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

Comparing EV/EV.pm (file contents):
Revision 1.26 by root, Fri Nov 2 23:22:17 2007 UTC vs.
Revision 1.52 by root, Sat Nov 24 16:20:30 2007 UTC

10 10
11 my $w = EV::timer 2, 0, sub { 11 my $w = EV::timer 2, 0, sub {
12 warn "is called after 2s"; 12 warn "is called after 2s";
13 }; 13 };
14 14
15 my $w = EV::timer 2, 1, sub { 15 my $w = EV::timer 2, 2, sub {
16 warn "is called roughly every 2s (repeat = 1)"; 16 warn "is called roughly every 2s (repeat = 2)";
17 }; 17 };
18 18
19 undef $w; # destroy event watcher again 19 undef $w; # destroy event watcher again
20 20
21 my $w = EV::periodic 0, 60, sub { 21 my $w = EV::periodic 0, 60, 0, sub {
22 warn "is called every minute, on the minute, exactly"; 22 warn "is called every minute, on the minute, exactly";
23 }; 23 };
24 24
25 # IO 25 # IO
26 26
27 my $w = EV::io *STDIN, EV::READ, sub { 27 my $w = EV::io *STDIN, EV::READ, sub {
28 my ($w, $revents) = @_; # all callbacks get the watcher object and event mask 28 my ($w, $revents) = @_; # all callbacks receive the watcher and event mask
29 warn "stdin is readable, you entered: ", <STDIN>; 29 warn "stdin is readable, you entered: ", <STDIN>;
30 }; 30 };
31 31
32 # SIGNALS 32 # SIGNALS
33 33
34 my $w = EV::signal 'QUIT', sub { 34 my $w = EV::signal 'QUIT', sub {
35 warn "sigquit received\n"; 35 warn "sigquit received\n";
36 }; 36 };
37 37
38 # CHILD/PID STATUS CHANGES
39
38 my $w = EV::signal 3, sub { 40 my $w = EV::child 666, sub {
39 warn "sigquit received (this is GNU/Linux, right?)\n"; 41 my ($w, $revents) = @_;
42 my $status = $w->rstatus;
40 }; 43 };
41
42 # CHILD/PID STATUS CHANGES
43
44 my $w = EV::child 666, sub {
45 my ($w, $revents, $status) = @_;
46 };
47 44
48 # MAINLOOP 45 # MAINLOOP
49 EV::loop; # loop until EV::loop_done is called 46 EV::loop; # loop until EV::unloop is called or all watchers stop
50 EV::loop EV::LOOP_ONESHOT; # block until at least one event could be handled 47 EV::loop EV::LOOP_ONESHOT; # block until at least one event could be handled
51 EV::loop EV::LOOP_NONBLOCK; # try to handle same events, but do not block 48 EV::loop EV::LOOP_NONBLOCK; # try to handle same events, but do not block
52 49
53=head1 DESCRIPTION 50=head1 DESCRIPTION
54 51
55This module provides an interface to libev 52This module provides an interface to libev
56(L<http://software.schmorp.de/pkg/libev.html>). 53(L<http://software.schmorp.de/pkg/libev.html>). While the documentation
54below is comprehensive, one might also consult the documentation of libev
55itself (L<http://cvs.schmorp.de/libev/ev.html>) for more subtle details on
56watcher semantics or some discussion on the available backends, or how to
57force a specific backend with C<LIBEV_FLAGS>.
57 58
58=cut 59=cut
59 60
60package EV; 61package EV;
61 62
62use strict; 63use strict;
63 64
64BEGIN { 65BEGIN {
65 our $VERSION = '0.5'; 66 our $VERSION = '1.3';
66 use XSLoader; 67 use XSLoader;
67 XSLoader::load "EV", $VERSION; 68 XSLoader::load "EV", $VERSION;
68} 69}
69 70
70@EV::Io::ISA = 71@EV::IO::ISA =
71@EV::Timer::ISA = 72@EV::Timer::ISA =
72@EV::Periodic::ISA = 73@EV::Periodic::ISA =
73@EV::Signal::ISA = 74@EV::Signal::ISA =
74@EV::Idle::ISA = 75@EV::Idle::ISA =
75@EV::Prepare::ISA = 76@EV::Prepare::ISA =
96 97
97Returns the time the last event loop iteration has been started. This 98Returns the time the last event loop iteration has been started. This
98is the time that (relative) timers are based on, and refering to it is 99is the time that (relative) timers are based on, and refering to it is
99usually faster then calling EV::time. 100usually faster then calling EV::time.
100 101
101=item $method = EV::ev_method 102=item $method = EV::method
102 103
103Returns an integer describing the backend used by libev (EV::METHOD_SELECT 104Returns an integer describing the backend used by libev (EV::METHOD_SELECT
104or EV::METHOD_EPOLL). 105or EV::METHOD_EPOLL).
105 106
106=item EV::loop [$flags] 107=item EV::loop [$flags]
107 108
108Begin checking for events and calling callbacks. It returns when a 109Begin checking for events and calling callbacks. It returns when a
109callback calls EV::loop_done. 110callback calls EV::unloop.
110 111
111The $flags argument can be one of the following: 112The $flags argument can be one of the following:
112 113
113 0 as above 114 0 as above
114 EV::LOOP_ONESHOT block at most once (wait, but do not loop) 115 EV::LOOP_ONESHOT block at most once (wait, but do not loop)
115 EV::LOOP_NONBLOCK do not block at all (fetch/handle events but do not wait) 116 EV::LOOP_NONBLOCK do not block at all (fetch/handle events but do not wait)
116 117
117=item EV::loop_done [$how] 118=item EV::unloop [$how]
118 119
119When called with no arguments or an argument of 1, makes the innermost 120When called with no arguments or an argument of EV::UNLOOP_ONE, makes the
120call to EV::loop return. 121innermost call to EV::loop return.
121 122
122When called with an agrument of 2, all calls to EV::loop will return as 123When called with an argument of EV::UNLOOP_ALL, all calls to EV::loop will return as
123fast as possible. 124fast as possible.
125
126=item EV::once $fh_or_undef, $events, $timeout, $cb->($revents)
127
128This function rolls together an I/O and a timer watcher for a single
129one-shot event without the need for managing a watcher object.
130
131If C<$fh_or_undef> is a filehandle or file descriptor, then C<$events>
132must be a bitset containing either C<EV::READ>, C<EV::WRITE> or C<EV::READ
133| EV::WRITE>, indicating the type of I/O event you want to wait for. If
134you do not want to wait for some I/O event, specify C<undef> for
135C<$fh_or_undef> and C<0> for C<$events>).
136
137If timeout is C<undef> or negative, then there will be no
138timeout. Otherwise a EV::timer with this value will be started.
139
140When an error occurs or either the timeout or I/O watcher triggers, then
141the callback will be called with the received event set (in general
142you can expect it to be a combination of C<EV:ERROR>, C<EV::READ>,
143C<EV::WRITE> and C<EV::TIMEOUT>).
144
145EV::once doesn't return anything: the watchers stay active till either
146of them triggers, then they will be stopped and freed, and the callback
147invoked.
124 148
125=back 149=back
126 150
127=head2 WATCHER 151=head2 WATCHER
128 152
184 208
185=item $bool = $w->is_active 209=item $bool = $w->is_active
186 210
187Returns true if the watcher is active, false otherwise. 211Returns true if the watcher is active, false otherwise.
188 212
213=item $current_data = $w->data
214
215=item $old_data = $w->data ($new_data)
216
217Queries a freely usable data scalar on the watcher and optionally changes
218it. This is a way to associate custom data with a watcher:
219
220 my $w = EV::timer 60, 0, sub {
221 warn $_[0]->data;
222 };
223 $w->data ("print me!");
224
189=item $current_cb = $w->cb 225=item $current_cb = $w->cb
190 226
191=item $old_cb = $w->cb ($new_cb) 227=item $old_cb = $w->cb ($new_cb)
192 228
193Queries the callback on the watcher and optionally changes it. You can do 229Queries the callback on the watcher and optionally changes it. You can do
201watchers with higher priority will be invoked first. The valid range of 237watchers with higher priority will be invoked first. The valid range of
202priorities lies between EV::MAXPRI (default 2) and EV::MINPRI (default 238priorities lies between EV::MAXPRI (default 2) and EV::MINPRI (default
203-2). If the priority is outside this range it will automatically be 239-2). If the priority is outside this range it will automatically be
204normalised to the nearest valid priority. 240normalised to the nearest valid priority.
205 241
206The default priority of any newly-created weatcher is 0. 242The default priority of any newly-created watcher is 0.
243
244Note that the priority semantics have not yet been fleshed out and are
245subject to almost certain change.
207 246
208=item $w->trigger ($revents) 247=item $w->trigger ($revents)
209 248
210Call the callback *now* with the given event mask. 249Call the callback *now* with the given event mask.
211 250
251=item $previous_state = $w->keepalive ($bool)
252
253Normally, C<EV::loop> will return when there are no active watchers
254(which is a "deadlock" because no progress can be made anymore). This is
255convinient because it allows you to start your watchers (and your jobs),
256call C<EV::loop> once and when it returns you know that all your jobs are
257finished (or they forgot to register some watchers for their task :).
258
259Sometimes, however, this gets in your way, for example when you the module
260that calls C<EV::loop> (usually the main program) is not the same module
261as a long-living watcher (for example a DNS client module written by
262somebody else even). Then you might want any outstanding requests to be
263handled, but you would not want to keep C<EV::loop> from returning just
264because you happen to have this long-running UDP port watcher.
265
266In this case you can clear the keepalive status, which means that even
267though your watcher is active, it won't keep C<EV::loop> from returning.
268
269The initial value for keepalive is true (enabled), and you cna change it
270any time.
271
272Example: Register an IO watcher for some UDP socket but do not keep the
273event loop from running just because of that watcher.
274
275 my $udp_socket = ...
276 my $udp_watcher = EV::io $udp_socket, EV::READ, sub { ... };
277 $udp_watcher->keepalive (0);
212 278
213=item $w = EV::io $fileno_or_fh, $eventmask, $callback 279=item $w = EV::io $fileno_or_fh, $eventmask, $callback
214 280
215=item $w = EV::io_ns $fileno_or_fh, $eventmask, $callback 281=item $w = EV::io_ns $fileno_or_fh, $eventmask, $callback
216 282
249Calls the callback after C<$after> seconds. If C<$repeat> is non-zero, 315Calls the callback after C<$after> seconds. If C<$repeat> is non-zero,
250the timer will be restarted (with the $repeat value as $after) after the 316the timer will be restarted (with the $repeat value as $after) after the
251callback returns. 317callback returns.
252 318
253This means that the callback would be called roughly after C<$after> 319This means that the callback would be called roughly after C<$after>
254seconds, and then every C<$repeat> seconds. "Roughly" because the time of 320seconds, and then every C<$repeat> seconds. The timer does his best not
255callback processing is not taken into account, so the timer will slowly 321to drift, but it will not invoke the timer more often then once per event
256drift. If that isn't acceptable, look at EV::periodic. 322loop iteration, and might drift in other cases. If that isn't acceptable,
323look at EV::periodic, which can provide long-term stable timers.
257 324
258The timer is based on a monotonic clock, that is if somebody is sitting 325The timer is based on a monotonic clock, that is, if somebody is sitting
259in front of the machine while the timer is running and changes the system 326in front of the machine while the timer is running and changes the system
260clock, the timer will nevertheless run (roughly) the same time. 327clock, the timer will nevertheless run (roughly) the same time.
261 328
262The C<timer_ns> variant doesn't start (activate) the newly created watcher. 329The C<timer_ns> variant doesn't start (activate) the newly created watcher.
263 330
268 335
269=item $w->again 336=item $w->again
270 337
271Similar to the C<start> method, but has special semantics for repeating timers: 338Similar to the C<start> method, but has special semantics for repeating timers:
272 339
340If the timer is active and non-repeating, it will be stopped.
341
273If the timer is active and repeating, reset the timeout to occur 342If the timer is active and repeating, reset the timeout to occur
274C<$repeat> seconds after now. 343C<$repeat> seconds after now.
275 344
276If the timer is active and non-repeating, it will be stopped.
277
278If the timer is in active and repeating, start it. 345If the timer is inactive and repeating, start it using the repeat value.
279 346
280Otherwise do nothing. 347Otherwise do nothing.
281 348
282This behaviour is useful when you have a timeout for some IO 349This behaviour is useful when you have a timeout for some IO
283operation. You create a timer object with the same value for C<$after> and 350operation. You create a timer object with the same value for C<$after> and
284C<$repeat>, and then, in the read/write watcher, run the C<again> method 351C<$repeat>, and then, in the read/write watcher, run the C<again> method
285on the timeout. 352on the timeout.
286 353
287 354
288=item $w = EV::periodic $at, $interval, $callback 355=item $w = EV::periodic $at, $interval, $reschedule_cb, $callback
289 356
290=item $w = EV::periodic_ns $at, $interval, $callback 357=item $w = EV::periodic_ns $at, $interval, $reschedule_cb, $callback
291 358
292Similar to EV::timer, but the time is given as an absolute point in time 359Similar to EV::timer, but is not based on relative timeouts but on
293(C<$at>), plus an optional C<$interval>. 360absolute times. Apart from creating "simple" timers that trigger "at" the
361specified time, it can also be used for non-drifting absolute timers and
362more complex, cron-like, setups that are not adversely affected by time
363jumps (i.e. when the system clock is changed by explicit date -s or other
364means such as ntpd). It is also the most complex watcher type in EV.
294 365
295If the C<$interval> is zero, then the callback will be called at the time 366It has three distinct "modes":
296C<$at> if that is in the future, or as soon as possible if it is in the
297past. It will not automatically repeat.
298 367
299If the C<$interval> is nonzero, then the watcher will always be scheduled 368=over 4
300to time out at the next C<$at + N * $interval> time.
301 369
302This can be used to schedule a callback to run at very regular intervals, 370=item * absolute timer ($interval = $reschedule_cb = 0)
303as long as the processing time is less then the interval (otherwise 371
304obviously events will be skipped). 372This time simply fires at the wallclock time C<$at> and doesn't repeat. It
373will not adjust when a time jump occurs, that is, if it is to be run
374at January 1st 2011 then it will run when the system time reaches or
375surpasses this time.
376
377=item * non-repeating interval timer ($interval > 0, $reschedule_cb = 0)
378
379In this mode the watcher will always be scheduled to time out at the
380next C<$at + N * $interval> time (for some integer N) and then repeat,
381regardless of any time jumps.
382
383This can be used to create timers that do not drift with respect to system
384time:
385
386 my $hourly = EV::periodic 0, 3600, 0, sub { print "once/hour\n" };
387
388That doesn't mean there will always be 3600 seconds in between triggers,
389but only that the the clalback will be called when the system time shows a
390full hour (UTC).
305 391
306Another way to think about it (for the mathematically inclined) is that 392Another way to think about it (for the mathematically inclined) is that
307EV::periodic will try to run the callback at the next possible time where 393EV::periodic will try to run the callback in this mode at the next
308C<$time = $at (mod $interval)>, regardless of any time jumps. 394possible time where C<$time = $at (mod $interval)>, regardless of any time
395jumps.
309 396
310This periodic timer is based on "wallclock time", that is, if the clock 397=item * manual reschedule mode ($reschedule_cb = coderef)
311changes (C<ntp>, C<date -s> etc.), then the timer will nevertheless run at 398
312the specified time. This means it will never drift (it might jitter, but 399In this mode $interval and $at are both being ignored. Instead, each
313it will not drift). 400time the periodic watcher gets scheduled, the reschedule callback
401($reschedule_cb) will be called with the watcher as first, and the current
402time as second argument.
403
404I<This callback MUST NOT stop or destroy this or any other periodic
405watcher, ever>. If you need to stop it, return 1e30 and stop it
406afterwards.
407
408It must return the next time to trigger, based on the passed time value
409(that is, the lowest time value larger than to the second argument). It
410will usually be called just before the callback will be triggered, but
411might be called at other times, too.
412
413This can be used to create very complex timers, such as a timer that
414triggers on each midnight, local time (actually 24 hours after the last
415midnight, to keep the example simple. If you know a way to do it correctly
416in about the same space (without requiring elaborate modules), drop me a
417note :):
418
419 my $daily = EV::periodic 0, 0, sub {
420 my ($w, $now) = @_;
421
422 use Time::Local ();
423 my (undef, undef, undef, $d, $m, $y) = localtime $now;
424 86400 + Time::Local::timelocal 0, 0, 0, $d, $m, $y
425 }, sub {
426 print "it's midnight or likely shortly after, now\n";
427 };
428
429=back
314 430
315The C<periodic_ns> variant doesn't start (activate) the newly created watcher. 431The C<periodic_ns> variant doesn't start (activate) the newly created watcher.
316 432
317=item $w->set ($at, $interval) 433=item $w->set ($at, $interval, $reschedule_cb)
318 434
319Reconfigures the watcher, see the constructor above for details. Can be at 435Reconfigures the watcher, see the constructor above for details. Can be at
320any time. 436any time.
437
438=item $w->again
439
440Simply stops and starts the watcher again.
321 441
322 442
323=item $w = EV::signal $signal, $callback 443=item $w = EV::signal $signal, $callback
324 444
325=item $w = EV::signal_ns $signal, $callback 445=item $w = EV::signal_ns $signal, $callback
356Call the callback when a status change for pid C<$pid> (or any pid 476Call the callback when a status change for pid C<$pid> (or any pid
357if C<$pid> is 0) has been received. More precisely: when the process 477if C<$pid> is 0) has been received. More precisely: when the process
358receives a SIGCHLD, EV will fetch the outstanding exit/wait status for all 478receives a SIGCHLD, EV will fetch the outstanding exit/wait status for all
359changed/zombie children and call the callback. 479changed/zombie children and call the callback.
360 480
361Unlike all other callbacks, this callback will be called with an 481You can access both status and pid by using the C<rstatus> and C<rpid>
362additional third argument which is the exit status. See the C<waitpid> 482methods on the watcher object.
363function for details.
364 483
365You can have as many pid watchers per pid as you want. 484You can have as many pid watchers per pid as you want.
366 485
367The C<child_ns> variant doesn't start (activate) the newly created watcher. 486The C<child_ns> variant doesn't start (activate) the newly created watcher.
368 487
374=item $current_pid = $w->pid 493=item $current_pid = $w->pid
375 494
376=item $old_pid = $w->pid ($new_pid) 495=item $old_pid = $w->pid ($new_pid)
377 496
378Returns the previously set process id and optionally set a new one. 497Returns the previously set process id and optionally set a new one.
498
499=item $exit_status = $w->rstatus
500
501Return the exit/wait status (as returned by waitpid, see the waitpid entry
502in perlfunc).
503
504=item $pid = $w->rpid
505
506Return the pid of the awaited child (useful when you have installed a
507watcher for all pids).
379 508
380 509
381=item $w = EV::idle $callback 510=item $w = EV::idle $callback
382 511
383=item $w = EV::idle_ns $callback 512=item $w = EV::idle_ns $callback
421 # do nothing unless active 550 # do nothing unless active
422 $dispatcher->{_event_queue_h} 551 $dispatcher->{_event_queue_h}
423 or return; 552 or return;
424 553
425 # make the dispatcher handle any outstanding stuff 554 # make the dispatcher handle any outstanding stuff
555 ... not shown
426 556
427 # create an IO watcher for each and every socket 557 # create an IO watcher for each and every socket
428 @snmp_watcher = ( 558 @snmp_watcher = (
429 (map { EV::io $_, EV::READ, sub { } } 559 (map { EV::io $_, EV::READ, sub { } }
430 keys %{ $dispatcher->{_descriptors} }), 560 keys %{ $dispatcher->{_descriptors} }),
561
562 EV::timer +($event->[Net::SNMP::Dispatcher::_ACTIVE]
563 ? $event->[Net::SNMP::Dispatcher::_TIME] - EV::now : 0),
564 0, sub { },
431 ); 565 );
432
433 # if there are any timeouts, also create a timer
434 push @snmp_watcher, EV::timer $event->[Net::SNMP::Dispatcher::_TIME] - EV::now, 0, sub { }
435 if $event->[Net::SNMP::Dispatcher::_ACTIVE];
436 }; 566 };
437 567
438The callbacks are irrelevant, the only purpose of those watchers is 568The callbacks are irrelevant (and are not even being called), the
439to wake up the process as soon as one of those events occurs (socket 569only purpose of those watchers is to wake up the process as soon as
440readable, or timer timed out). The corresponding EV::check watcher will then 570one of those events occurs (socket readable, or timer timed out). The
441clean up: 571corresponding EV::check watcher will then clean up:
442 572
443 our $snmp_check = EV::check sub { 573 our $snmp_check = EV::check sub {
444 # destroy all watchers 574 # destroy all watchers
445 @snmp_watcher = (); 575 @snmp_watcher = ();
446 576
447 # make the dispatcher handle any new stuff 577 # make the dispatcher handle any new stuff
578 ... not shown
448 }; 579 };
449 580
450The callbacks of the created watchers will not be called as the watchers 581The callbacks of the created watchers will not be called as the watchers
451are destroyed before this cna happen (remember EV::check gets called 582are destroyed before this cna happen (remember EV::check gets called
452first). 583first).
455 586
456=back 587=back
457 588
458=head1 THREADS 589=head1 THREADS
459 590
460Threads are not supported by this in any way. Perl pseudo-threads is evil 591Threads are not supported by this module in any way. Perl pseudo-threads
461stuff and must die. 592is evil stuff and must die. As soon as Perl gains real threads I will work
593on thread support for it.
594
595=head1 FORK
596
597Most of the "improved" event delivering mechanisms of modern operating
598systems have quite a few problems with fork(2) (to put it bluntly: it is
599not supported and usually destructive). Libev makes it possible to work
600around this by having a function that recreates the kernel state after
601fork in the child.
602
603On non-win32 platforms, this module requires the pthread_atfork
604functionality to do this automatically for you. This function is quite
605buggy on most BSDs, though, so YMMV. The overhead for this is quite
606negligible, because everything the function currently does is set a flag
607that is checked only when the event loop gets used the next time, so when
608you do fork but not use EV, the overhead is minimal.
609
610On win32, there is no notion of fork so all this doesn't apply, of course.
462 611
463=cut 612=cut
464 613
465our $DIED = sub { 614our $DIED = sub {
466 warn "EV: error in callback (ignoring): $@"; 615 warn "EV: error in callback (ignoring): $@";
467}; 616};
468 617
469init 618default_loop
470 or die 'EV: cannot initialise libev backend. bad $ENV{LIBEV_METHODS}?'; 619 or die 'EV: cannot initialise libev backend. bad $ENV{LIBEV_METHODS}?';
471 620
472push @AnyEvent::REGISTRY, [EV => "EV::AnyEvent"];
473
4741; 6211;
475 622
476=head1 SEE ALSO 623=head1 SEE ALSO
477 624
478 L<EV::DNS>, L<EV::AnyEvent>. 625 L<EV::DNS>.
479 626
480=head1 AUTHOR 627=head1 AUTHOR
481 628
482 Marc Lehmann <schmorp@schmorp.de> 629 Marc Lehmann <schmorp@schmorp.de>
483 http://home.schmorp.de/ 630 http://home.schmorp.de/

Diff Legend

Removed lines
+ Added lines
< Changed lines
> Changed lines