| 1 |
NAME |
| 2 |
EV - perl interface to libev, a high performance full-featured event |
| 3 |
loop |
| 4 |
|
| 5 |
SYNOPSIS |
| 6 |
use EV; |
| 7 |
|
| 8 |
# TIMERS |
| 9 |
|
| 10 |
my $w = EV::timer 2, 0, sub { |
| 11 |
warn "is called after 2s"; |
| 12 |
}; |
| 13 |
|
| 14 |
my $w = EV::timer 2, 2, sub { |
| 15 |
warn "is called roughly every 2s (repeat = 2)"; |
| 16 |
}; |
| 17 |
|
| 18 |
undef $w; # destroy event watcher again |
| 19 |
|
| 20 |
my $w = EV::periodic 0, 60, 0, sub { |
| 21 |
warn "is called every minute, on the minute, exactly"; |
| 22 |
}; |
| 23 |
|
| 24 |
# IO |
| 25 |
|
| 26 |
my $w = EV::io *STDIN, EV::READ, sub { |
| 27 |
my ($w, $revents) = @_; # all callbacks receive the watcher and event mask |
| 28 |
warn "stdin is readable, you entered: ", <STDIN>; |
| 29 |
}; |
| 30 |
|
| 31 |
# SIGNALS |
| 32 |
|
| 33 |
my $w = EV::signal 'QUIT', sub { |
| 34 |
warn "sigquit received\n"; |
| 35 |
}; |
| 36 |
|
| 37 |
# CHILD/PID STATUS CHANGES |
| 38 |
|
| 39 |
my $w = EV::child 666, sub { |
| 40 |
my ($w, $revents) = @_; |
| 41 |
my $status = $w->rstatus; |
| 42 |
}; |
| 43 |
|
| 44 |
# MAINLOOP |
| 45 |
EV::loop; # loop until EV::unloop is called or all watchers stop |
| 46 |
EV::loop EV::LOOP_ONESHOT; # block until at least one event could be handled |
| 47 |
EV::loop EV::LOOP_NONBLOCK; # try to handle same events, but do not block |
| 48 |
|
| 49 |
DESCRIPTION |
| 50 |
This module provides an interface to libev |
| 51 |
(<http://software.schmorp.de/pkg/libev.html>). |
| 52 |
|
| 53 |
BASIC INTERFACE |
| 54 |
$EV::DIED |
| 55 |
Must contain a reference to a function that is called when a |
| 56 |
callback throws an exception (with $@ containing thr error). The |
| 57 |
default prints an informative message and continues. |
| 58 |
|
| 59 |
If this callback throws an exception it will be silently ignored. |
| 60 |
|
| 61 |
$time = EV::time |
| 62 |
Returns the current time in (fractional) seconds since the epoch. |
| 63 |
|
| 64 |
$time = EV::now |
| 65 |
Returns the time the last event loop iteration has been started. |
| 66 |
This is the time that (relative) timers are based on, and refering |
| 67 |
to it is usually faster then calling EV::time. |
| 68 |
|
| 69 |
$method = EV::ev_method |
| 70 |
Returns an integer describing the backend used by libev |
| 71 |
(EV::METHOD_SELECT or EV::METHOD_EPOLL). |
| 72 |
|
| 73 |
EV::loop [$flags] |
| 74 |
Begin checking for events and calling callbacks. It returns when a |
| 75 |
callback calls EV::unloop. |
| 76 |
|
| 77 |
The $flags argument can be one of the following: |
| 78 |
|
| 79 |
0 as above |
| 80 |
EV::LOOP_ONESHOT block at most once (wait, but do not loop) |
| 81 |
EV::LOOP_NONBLOCK do not block at all (fetch/handle events but do not wait) |
| 82 |
|
| 83 |
EV::unloop [$how] |
| 84 |
When called with no arguments or an argument of EV::UNLOOP_ONE, |
| 85 |
makes the innermost call to EV::loop return. |
| 86 |
|
| 87 |
When called with an argument of EV::UNLOOP_ALL, all calls to |
| 88 |
EV::loop will return as fast as possible. |
| 89 |
|
| 90 |
WATCHER |
| 91 |
A watcher is an object that gets created to record your interest in some |
| 92 |
event. For instance, if you want to wait for STDIN to become readable, |
| 93 |
you would create an EV::io watcher for that: |
| 94 |
|
| 95 |
my $watcher = EV::io *STDIN, EV::READ, sub { |
| 96 |
my ($watcher, $revents) = @_; |
| 97 |
warn "yeah, STDIN should not be readable without blocking!\n" |
| 98 |
}; |
| 99 |
|
| 100 |
All watchers can be active (waiting for events) or inactive (paused). |
| 101 |
Only active watchers will have their callbacks invoked. All callbacks |
| 102 |
will be called with at least two arguments: the watcher and a bitmask of |
| 103 |
received events. |
| 104 |
|
| 105 |
Each watcher type has its associated bit in revents, so you can use the |
| 106 |
same callback for multiple watchers. The event mask is named after the |
| 107 |
type, i..e. EV::child sets EV::CHILD, EV::prepare sets EV::PREPARE, |
| 108 |
EV::periodic sets EV::PERIODIC and so on, with the exception of IO |
| 109 |
events (which can set both EV::READ and EV::WRITE bits), and EV::timer |
| 110 |
(which uses EV::TIMEOUT). |
| 111 |
|
| 112 |
In the rare case where one wants to create a watcher but not start it at |
| 113 |
the same time, each constructor has a variant with a trailing "_ns" in |
| 114 |
its name, e.g. EV::io has a non-starting variant EV::io_ns and so on. |
| 115 |
|
| 116 |
Please note that a watcher will automatically be stopped when the |
| 117 |
watcher object is destroyed, so you *need* to keep the watcher objects |
| 118 |
returned by the constructors. |
| 119 |
|
| 120 |
Also, all methods changing some aspect of a watcher (->set, ->priority, |
| 121 |
->fh and so on) automatically stop and start it again if it is active, |
| 122 |
which means pending events get lost. |
| 123 |
|
| 124 |
WATCHER TYPES |
| 125 |
Now lets move to the existing watcher types and asociated methods. |
| 126 |
|
| 127 |
The following methods are available for all watchers. Then followes a |
| 128 |
description of each watcher constructor (EV::io, EV::timer, |
| 129 |
EV::periodic, EV::signal, EV::child, EV::idle, EV::prepare and |
| 130 |
EV::check), followed by any type-specific methods (if any). |
| 131 |
|
| 132 |
$w->start |
| 133 |
Starts a watcher if it isn't active already. Does nothing to an |
| 134 |
already active watcher. By default, all watchers start out in the |
| 135 |
active state (see the description of the "_ns" variants if you need |
| 136 |
stopped watchers). |
| 137 |
|
| 138 |
$w->stop |
| 139 |
Stop a watcher if it is active. Also clear any pending events |
| 140 |
(events that have been received but that didn't yet result in a |
| 141 |
callback invocation), regardless of wether the watcher was active or |
| 142 |
not. |
| 143 |
|
| 144 |
$bool = $w->is_active |
| 145 |
Returns true if the watcher is active, false otherwise. |
| 146 |
|
| 147 |
$current_data = $w->data |
| 148 |
$old_data = $w->data ($new_data) |
| 149 |
Queries a freely usable data scalar on the watcher and optionally |
| 150 |
changes it. This is a way to associate custom data with a watcher: |
| 151 |
|
| 152 |
my $w = EV::timer 60, 0, sub { |
| 153 |
warn $_[0]->data; |
| 154 |
}; |
| 155 |
$w->data ("print me!"); |
| 156 |
|
| 157 |
$current_cb = $w->cb |
| 158 |
$old_cb = $w->cb ($new_cb) |
| 159 |
Queries the callback on the watcher and optionally changes it. You |
| 160 |
can do this at any time without the watcher restarting. |
| 161 |
|
| 162 |
$current_priority = $w->priority |
| 163 |
$old_priority = $w->priority ($new_priority) |
| 164 |
Queries the priority on the watcher and optionally changes it. |
| 165 |
Pending watchers with higher priority will be invoked first. The |
| 166 |
valid range of priorities lies between EV::MAXPRI (default 2) and |
| 167 |
EV::MINPRI (default -2). If the priority is outside this range it |
| 168 |
will automatically be normalised to the nearest valid priority. |
| 169 |
|
| 170 |
The default priority of any newly-created weatcher is 0. |
| 171 |
|
| 172 |
$w->trigger ($revents) |
| 173 |
Call the callback *now* with the given event mask. |
| 174 |
|
| 175 |
$w = EV::io $fileno_or_fh, $eventmask, $callback |
| 176 |
$w = EV::io_ns $fileno_or_fh, $eventmask, $callback |
| 177 |
As long as the returned watcher object is alive, call the $callback |
| 178 |
when the events specified in $eventmask. |
| 179 |
|
| 180 |
The $eventmask can be one or more of these constants ORed together: |
| 181 |
|
| 182 |
EV::READ wait until read() wouldn't block anymore |
| 183 |
EV::WRITE wait until write() wouldn't block anymore |
| 184 |
|
| 185 |
The "io_ns" variant doesn't start (activate) the newly created |
| 186 |
watcher. |
| 187 |
|
| 188 |
$w->set ($fileno_or_fh, $eventmask) |
| 189 |
Reconfigures the watcher, see the constructor above for details. Can |
| 190 |
be called at any time. |
| 191 |
|
| 192 |
$current_fh = $w->fh |
| 193 |
$old_fh = $w->fh ($new_fh) |
| 194 |
Returns the previously set filehandle and optionally set a new one. |
| 195 |
|
| 196 |
$current_eventmask = $w->events |
| 197 |
$old_eventmask = $w->events ($new_eventmask) |
| 198 |
Returns the previously set event mask and optionally set a new one. |
| 199 |
|
| 200 |
$w = EV::timer $after, $repeat, $callback |
| 201 |
$w = EV::timer_ns $after, $repeat, $callback |
| 202 |
Calls the callback after $after seconds. If $repeat is non-zero, the |
| 203 |
timer will be restarted (with the $repeat value as $after) after the |
| 204 |
callback returns. |
| 205 |
|
| 206 |
This means that the callback would be called roughly after $after |
| 207 |
seconds, and then every $repeat seconds. The timer does his best not |
| 208 |
to drift, but it will not invoke the timer more often then once per |
| 209 |
event loop iteration, and might drift in other cases. If that isn't |
| 210 |
acceptable, look at EV::periodic, which can provide long-term stable |
| 211 |
timers. |
| 212 |
|
| 213 |
The timer is based on a monotonic clock, that is, if somebody is |
| 214 |
sitting in front of the machine while the timer is running and |
| 215 |
changes the system clock, the timer will nevertheless run (roughly) |
| 216 |
the same time. |
| 217 |
|
| 218 |
The "timer_ns" variant doesn't start (activate) the newly created |
| 219 |
watcher. |
| 220 |
|
| 221 |
$w->set ($after, $repeat) |
| 222 |
Reconfigures the watcher, see the constructor above for details. Can |
| 223 |
be at any time. |
| 224 |
|
| 225 |
$w->again |
| 226 |
Similar to the "start" method, but has special semantics for |
| 227 |
repeating timers: |
| 228 |
|
| 229 |
If the timer is active and non-repeating, it will be stopped. |
| 230 |
|
| 231 |
If the timer is active and repeating, reset the timeout to occur |
| 232 |
$repeat seconds after now. |
| 233 |
|
| 234 |
If the timer is inactive and repeating, start it using the repeat |
| 235 |
value. |
| 236 |
|
| 237 |
Otherwise do nothing. |
| 238 |
|
| 239 |
This behaviour is useful when you have a timeout for some IO |
| 240 |
operation. You create a timer object with the same value for $after |
| 241 |
and $repeat, and then, in the read/write watcher, run the "again" |
| 242 |
method on the timeout. |
| 243 |
|
| 244 |
$w = EV::periodic $at, $interval, $reschedule_cb, $callback |
| 245 |
$w = EV::periodic_ns $at, $interval, $reschedule_cb, $callback |
| 246 |
Similar to EV::timer, but is not based on relative timeouts but on |
| 247 |
absolute times. Apart from creating "simple" timers that trigger |
| 248 |
"at" the specified time, it can also be used for non-drifting |
| 249 |
absolute timers and more complex, cron-like, setups that are not |
| 250 |
adversely affected by time jumps (i.e. when the system clock is |
| 251 |
changed by explicit date -s or other means such as ntpd). It is also |
| 252 |
the most complex watcher type in EV. |
| 253 |
|
| 254 |
It has three distinct "modes": |
| 255 |
|
| 256 |
* absolute timer ($interval = $reschedule_cb = 0) |
| 257 |
This time simply fires at the wallclock time $at and doesn't |
| 258 |
repeat. It will not adjust when a time jump occurs, that is, if |
| 259 |
it is to be run at January 1st 2011 then it will run when the |
| 260 |
system time reaches or surpasses this time. |
| 261 |
|
| 262 |
* non-repeating interval timer ($interval > 0, $reschedule_cb = 0) |
| 263 |
In this mode the watcher will always be scheduled to time out at |
| 264 |
the next "$at + N * $interval" time (for some integer N) and |
| 265 |
then repeat, regardless of any time jumps. |
| 266 |
|
| 267 |
This can be used to create timers that do not drift with respect |
| 268 |
to system time: |
| 269 |
|
| 270 |
my $hourly = EV::periodic 0, 3600, 0, sub { print "once/hour\n" }; |
| 271 |
|
| 272 |
That doesn't mean there will always be 3600 seconds in between |
| 273 |
triggers, but only that the the clalback will be called when the |
| 274 |
system time shows a full hour (UTC). |
| 275 |
|
| 276 |
Another way to think about it (for the mathematically inclined) |
| 277 |
is that EV::periodic will try to run the callback in this mode |
| 278 |
at the next possible time where "$time = $at (mod $interval)", |
| 279 |
regardless of any time jumps. |
| 280 |
|
| 281 |
* manual reschedule mode ($reschedule_cb = coderef) |
| 282 |
In this mode $interval and $at are both being ignored. Instead, |
| 283 |
each time the periodic watcher gets scheduled, the reschedule |
| 284 |
callback ($reschedule_cb) will be called with the watcher as |
| 285 |
first, and the current time as second argument. |
| 286 |
|
| 287 |
*This callback MUST NOT stop or destroy this or any other |
| 288 |
periodic watcher, ever*. If you need to stop it, return 1e30 and |
| 289 |
stop it afterwards. |
| 290 |
|
| 291 |
It must return the next time to trigger, based on the passed |
| 292 |
time value (that is, the lowest time value larger than to the |
| 293 |
second argument). It will usually be called just before the |
| 294 |
callback will be triggered, but might be called at other times, |
| 295 |
too. |
| 296 |
|
| 297 |
This can be used to create very complex timers, such as a timer |
| 298 |
that triggers on each midnight, local time (actually 24 hours |
| 299 |
after the last midnight, to keep the example simple. If you know |
| 300 |
a way to do it correctly in about the same space (without |
| 301 |
requiring elaborate modules), drop me a note :): |
| 302 |
|
| 303 |
my $daily = EV::periodic 0, 0, sub { |
| 304 |
my ($w, $now) = @_; |
| 305 |
|
| 306 |
use Time::Local (); |
| 307 |
my (undef, undef, undef, $d, $m, $y) = localtime $now; |
| 308 |
86400 + Time::Local::timelocal 0, 0, 0, $d, $m, $y |
| 309 |
}, sub { |
| 310 |
print "it's midnight or likely shortly after, now\n"; |
| 311 |
}; |
| 312 |
|
| 313 |
The "periodic_ns" variant doesn't start (activate) the newly created |
| 314 |
watcher. |
| 315 |
|
| 316 |
$w->set ($at, $interval, $reschedule_cb) |
| 317 |
Reconfigures the watcher, see the constructor above for details. Can |
| 318 |
be at any time. |
| 319 |
|
| 320 |
$w->again |
| 321 |
Simply stops and starts the watcher again. |
| 322 |
|
| 323 |
$w = EV::signal $signal, $callback |
| 324 |
$w = EV::signal_ns $signal, $callback |
| 325 |
Call the callback when $signal is received (the signal can be |
| 326 |
specified by number or by name, just as with kill or %SIG). |
| 327 |
|
| 328 |
EV will grab the signal for the process (the kernel only allows one |
| 329 |
component to receive a signal at a time) when you start a signal |
| 330 |
watcher, and removes it again when you stop it. Perl does the same |
| 331 |
when you add/remove callbacks to %SIG, so watch out. |
| 332 |
|
| 333 |
You can have as many signal watchers per signal as you want. |
| 334 |
|
| 335 |
The "signal_ns" variant doesn't start (activate) the newly created |
| 336 |
watcher. |
| 337 |
|
| 338 |
$w->set ($signal) |
| 339 |
Reconfigures the watcher, see the constructor above for details. Can |
| 340 |
be at any time. |
| 341 |
|
| 342 |
$current_signum = $w->signal |
| 343 |
$old_signum = $w->signal ($new_signal) |
| 344 |
Returns the previously set signal (always as a number not name) and |
| 345 |
optionally set a new one. |
| 346 |
|
| 347 |
$w = EV::child $pid, $callback |
| 348 |
$w = EV::child_ns $pid, $callback |
| 349 |
Call the callback when a status change for pid $pid (or any pid if |
| 350 |
$pid is 0) has been received. More precisely: when the process |
| 351 |
receives a SIGCHLD, EV will fetch the outstanding exit/wait status |
| 352 |
for all changed/zombie children and call the callback. |
| 353 |
|
| 354 |
You can access both status and pid by using the "rstatus" and "rpid" |
| 355 |
methods on the watcher object. |
| 356 |
|
| 357 |
You can have as many pid watchers per pid as you want. |
| 358 |
|
| 359 |
The "child_ns" variant doesn't start (activate) the newly created |
| 360 |
watcher. |
| 361 |
|
| 362 |
$w->set ($pid) |
| 363 |
Reconfigures the watcher, see the constructor above for details. Can |
| 364 |
be at any time. |
| 365 |
|
| 366 |
$current_pid = $w->pid |
| 367 |
$old_pid = $w->pid ($new_pid) |
| 368 |
Returns the previously set process id and optionally set a new one. |
| 369 |
|
| 370 |
$exit_status = $w->rstatus |
| 371 |
Return the exit/wait status (as returned by waitpid, see the waitpid |
| 372 |
entry in perlfunc). |
| 373 |
|
| 374 |
$pid = $w->rpid |
| 375 |
Return the pid of the awaited child (useful when you have installed |
| 376 |
a watcher for all pids). |
| 377 |
|
| 378 |
$w = EV::idle $callback |
| 379 |
$w = EV::idle_ns $callback |
| 380 |
Call the callback when there are no pending io, timer/periodic, |
| 381 |
signal or child events, i.e. when the process is idle. |
| 382 |
|
| 383 |
The process will not block as long as any idle watchers are active, |
| 384 |
and they will be called repeatedly until stopped. |
| 385 |
|
| 386 |
The "idle_ns" variant doesn't start (activate) the newly created |
| 387 |
watcher. |
| 388 |
|
| 389 |
$w = EV::prepare $callback |
| 390 |
$w = EV::prepare_ns $callback |
| 391 |
Call the callback just before the process would block. You can still |
| 392 |
create/modify any watchers at this point. |
| 393 |
|
| 394 |
See the EV::check watcher, below, for explanations and an example. |
| 395 |
|
| 396 |
The "prepare_ns" variant doesn't start (activate) the newly created |
| 397 |
watcher. |
| 398 |
|
| 399 |
$w = EV::check $callback |
| 400 |
$w = EV::check_ns $callback |
| 401 |
Call the callback just after the process wakes up again (after it |
| 402 |
has gathered events), but before any other callbacks have been |
| 403 |
invoked. |
| 404 |
|
| 405 |
This is used to integrate other event-based software into the EV |
| 406 |
mainloop: You register a prepare callback and in there, you create |
| 407 |
io and timer watchers as required by the other software. Here is a |
| 408 |
real-world example of integrating Net::SNMP (with some details left |
| 409 |
out): |
| 410 |
|
| 411 |
our @snmp_watcher; |
| 412 |
|
| 413 |
our $snmp_prepare = EV::prepare sub { |
| 414 |
# do nothing unless active |
| 415 |
$dispatcher->{_event_queue_h} |
| 416 |
or return; |
| 417 |
|
| 418 |
# make the dispatcher handle any outstanding stuff |
| 419 |
|
| 420 |
# create an IO watcher for each and every socket |
| 421 |
@snmp_watcher = ( |
| 422 |
(map { EV::io $_, EV::READ, sub { } } |
| 423 |
keys %{ $dispatcher->{_descriptors} }), |
| 424 |
); |
| 425 |
|
| 426 |
# if there are any timeouts, also create a timer |
| 427 |
push @snmp_watcher, EV::timer $event->[Net::SNMP::Dispatcher::_TIME] - EV::now, 0, sub { } |
| 428 |
if $event->[Net::SNMP::Dispatcher::_ACTIVE]; |
| 429 |
}; |
| 430 |
|
| 431 |
The callbacks are irrelevant, the only purpose of those watchers is |
| 432 |
to wake up the process as soon as one of those events occurs (socket |
| 433 |
readable, or timer timed out). The corresponding EV::check watcher |
| 434 |
will then clean up: |
| 435 |
|
| 436 |
our $snmp_check = EV::check sub { |
| 437 |
# destroy all watchers |
| 438 |
@snmp_watcher = (); |
| 439 |
|
| 440 |
# make the dispatcher handle any new stuff |
| 441 |
}; |
| 442 |
|
| 443 |
The callbacks of the created watchers will not be called as the |
| 444 |
watchers are destroyed before this cna happen (remember EV::check |
| 445 |
gets called first). |
| 446 |
|
| 447 |
The "check_ns" variant doesn't start (activate) the newly created |
| 448 |
watcher. |
| 449 |
|
| 450 |
THREADS |
| 451 |
Threads are not supported by this in any way. Perl pseudo-threads is |
| 452 |
evil stuff and must die. |
| 453 |
|
| 454 |
SEE ALSO |
| 455 |
L<EV::DNS>, L<EV::AnyEvent>. |
| 456 |
|
| 457 |
AUTHOR |
| 458 |
Marc Lehmann <schmorp@schmorp.de> |
| 459 |
http://home.schmorp.de/ |
| 460 |
|
