AnyEvent/lib/AnyEvent.pm

=head1 NAME

AnyEvent - provide framework for multiple event loops

Event, Coro, Glib, Tk - various supported event loops

=head1 SYNOPSIS

use AnyEvent;

   my $w = AnyEvent->io (fh => ..., poll => "[rw]+", cb => sub {
      my ($poll_got) = @_;
      ...
   });

- only one io watcher per $fh and $poll type is allowed
(i.e. on a socket you can have one r + one w or one rw
watcher, not any more.

- AnyEvent will keep filehandles alive, so as long as the watcher exists,
the filehandle exists.

   my $w = AnyEvent->timer (after => $seconds, cb => sub {
      ...
   });

- io and time watchers get canceled whenever $w is destroyed, so keep a copy

- timers can only be used once and must be recreated for repeated operation

   my $w = AnyEvent->condvar; # kind of main loop replacement
   $w->wait; # enters main loop till $condvar gets ->broadcast
   $w->broadcast; # wake up current and all future wait's

- condvars are used to give blocking behaviour when neccessary. Create
a condvar for any "request" or "event" your module might create, C<<
->broadcast >> it when the event happens and provide a function that calls
C<< ->wait >> for it. See the examples below.

=head1 DESCRIPTION

L<AnyEvent> provides an identical interface to multiple event loops. This
allows module authors to utilizy an event loop without forcing module
users to use the same event loop (as only a single event loop can coexist
peacefully at any one time).

The interface itself is vaguely similar but not identical to the Event
module.

On the first call of any method, the module tries to detect the currently
loaded event loop by probing wether any of the following modules is
loaded: L<Coro::Event>, L<Event>, L<Glib>, L<Tk>. The first one found is
used. If none is found, the module tries to load these modules in the
order given. The first one that could be successfully loaded will be
used. If still none could be found, it will issue an error.

=over 4

=cut

package AnyEvent;

no warnings;
use strict 'vars';
use Carp;

our $VERSION = 0.3;
our $MODEL;

our $AUTOLOAD;
our @ISA;

my @models = (
      [Coro  => Coro::Event::],
      [Event => Event::],
      [Glib  => Glib::],
      [Tk    => Tk::],
);

our %method = map +($_ => 1), qw(io timer condvar broadcast wait cancel DESTROY);

sub AUTOLOAD {
   $AUTOLOAD =~ s/.*://;

   $method{$AUTOLOAD}
      or croak "$AUTOLOAD: not a valid method for AnyEvent objects";

   unless ($MODEL) {
      # check for already loaded models
      for (@models) {
         my ($model, $package) = @$_;
         if (scalar keys %{ *{"$package\::"} }) {
            eval "require AnyEvent::Impl::$model";
            last if $MODEL;
         }
      }

      unless ($MODEL) {
         # try to load a model

         for (@models) {
            my ($model, $package) = @$_;
            eval "require AnyEvent::Impl::$model";
            last if $MODEL;
         }

         $MODEL
           or die "No event module selected for AnyEvent and autodetect failed. Install any one of these modules: Coro, Event, Glib or Tk.";
      }
   }

   @ISA = $MODEL;

   my $class = shift;
   $class->$AUTOLOAD (@_);
}

=back

=head1 EXAMPLE

The following program uses an io watcher to read data from stdin, a timer
to display a message once per second, and a condvar to exit the program
when the user enters quit:

   use AnyEvent;

   my $cv = AnyEvent->condvar;

   my $io_watcher = AnyEvent->io (fh => \*STDIN, poll => 'r', cb => sub {
      warn "io event <$_[0]>\n";   # will always output <r>
      chomp (my $input = <STDIN>); # read a line
      warn "read: $input\n";       # output what has been read
      $cv->broadcast if $input =~ /^q/i; # quit program if /^q/i
   });

   my $time_watcher; # can only be used once

   sub new_timer {
      $timer = AnyEvent->timer (after => 1, cb => sub {
         warn "timeout\n"; # print 'timeout' about every second
         &new_timer; # and restart the time
      });
   }

   new_timer; # create first timer

   $cv->wait; # wait until user enters /^q/i

=head1 REAL-WORLD EXAMPLE

Consider the L<Net::FCP> module. It features (among others) the following
API calls, which are to freenet what HTTP GET requests are to http:

   my $data = $fcp->client_get ($url); # blocks

   my $transaction = $fcp->txn_client_get ($url); # does not block
   $transaction->cb ( sub { ... } ); # set optional result callback
   my $data = $transaction->result; # possibly blocks

The C<client_get> method works like C<LWP::Simple::get>: it requests the
given URL and waits till the data has arrived. It is defined to be:

   sub client_get { $_[0]->txn_client_get ($_[1])->result }

And in fact is automatically generated. This is the blocking API of
L<Net::FCP>, and it works as simple as in any other, similar, module.

More complicated is C<txn_client_get>: It only creates a transaction
(completion, result, ...) object and initiates the transaction.

   my $txn = bless { }, Net::FCP::Txn::;

It also creates a condition variable that is used to signal the completion
of the request:

   $txn->{finished} = AnyAvent->condvar;

It then creates a socket in non-blocking mode.

   socket $txn->{fh}, ...;
   fcntl $txn->{fh}, F_SETFL, O_NONBLOCK;
   connect $txn->{fh}, ...
      and !$!{EWOULDBLOCK}
      and !$!{EINPROGRESS}
      and Carp::croak "unable to connect: $!\n";

Then it creates a write-watcher which gets called whenever an error occurs
or the connection succeeds:

   $txn->{w} = AnyEvent->io (fh => $txn->{fh}, poll => 'w', cb => sub { $txn->fh_ready_w });

And returns this transaction object. The C<fh_ready_w> callback gets
called as soon as the event loop detects that the socket is ready for
writing.

The C<fh_ready_w> method makes the socket blocking again, writes the
request data and replaces the watcher by a read watcher (waiting for reply
data). The actual code is more complicated, but that doesn't matter for
this example:

   fcntl $txn->{fh}, F_SETFL, 0;
   syswrite $txn->{fh}, $txn->{request}
      or die "connection or write error";
   $txn->{w} = AnyEvent->io (fh => $txn->{fh}, poll => 'r', cb => sub { $txn->fh_ready_r });

Again, C<fh_ready_r> waits till all data has arrived, and then stores the
result and signals any possible waiters that the request ahs finished:

   sysread $txn->{fh}, $txn->{buf}, length $txn->{$buf};

   if (end-of-file or data complete) {
     $txn->{result} = $txn->{buf};
     $txn->{finished}->broadcast;
     $txb->{cb}->($txn) of $txn->{cb}; # also call callback
   }

The C<result> method, finally, just waits for the finished signal (if the
request was already finished, it doesn't wait, of course, and returns the
data:

   $txn->{finished}->wait;
   return $txn->{result};

The actual code goes further and collects all errors (C<die>s, exceptions)
that occured during request processing. The C<result> method detects
wether an exception as thrown (it is stored inside the $txn object)
and just throws the exception, which means connection errors and other
problems get reported tot he code that tries to use the result, not in a
random callback.

All of this enables the following usage styles:

1. Blocking:

   my $data = $fcp->client_get ($url);

2. Blocking, but parallelizing:

   my @datas = map $_->result,
                  map $fcp->txn_client_get ($_),
                     @urls;

Both blocking examples work without the module user having to know
anything about events.

3a. Event-based in a main program, using any support Event module:

   use Event;

   $fcp->txn_client_get ($url)->cb (sub {
      my $txn = shift;
      my $data = $txn->result;
      ...
   });

   Event::loop;

3b. The module user could use AnyEvent, too:

   use AnyEvent;

   my $quit = AnyEvent->condvar;

   $fcp->txn_client_get ($url)->cb (sub {
      ...
      $quit->broadcast;
   });

   $quit->wait;

=head1 SEE ALSO

Event modules: L<Coro::Event>, L<Coro>, L<Event>, L<Glib::Event>, L<Glib>.

Implementations: L<AnyEvent::Impl::Coro>, L<AnyEvent::Impl::Event>, L<AnyEvent::Impl::Glib>, L<AnyEvent::Impl::Tk>.

Nontrivial usage example: L<Net::FCP>.

=head1

=cut

1

Revision:	1.6
Committed:	Mon Dec 19 17:03:29 2005 UTC (18 years, 6 months ago) by root
Branch:	MAIN
Changes since 1.5:	+3 -2 lines
Log Message:	* empty log message *
#	User	Rev	Content
1	root	1.1	=head1 NAME
2
3	root	1.2	AnyEvent - provide framework for multiple event loops
4
5			Event, Coro, Glib, Tk - various supported event loops
6	root	1.1
7			=head1 SYNOPSIS
8
9	root	1.2	use AnyEvent;
10
11	root	1.5	my $w = AnyEvent->io (fh => ..., poll => "[rw]+", cb => sub {
12	root	1.2	my ($poll_got) = @_;
13			...
14			});
15	root	1.5
16			- only one io watcher per $fh and $poll type is allowed
17			(i.e. on a socket you can have one r + one w or one rw
18			watcher, not any more.
19
20			- AnyEvent will keep filehandles alive, so as long as the watcher exists,
21			the filehandle exists.
22
23			my $w = AnyEvent->timer (after => $seconds, cb => sub {
24	root	1.2	...
25			});
26
27	root	1.5	- io and time watchers get canceled whenever $w is destroyed, so keep a copy
28
29			- timers can only be used once and must be recreated for repeated operation
30	root	1.2
31			my $w = AnyEvent->condvar; # kind of main loop replacement
32	root	1.5	$w->wait; # enters main loop till $condvar gets ->broadcast
33			$w->broadcast; # wake up current and all future wait's
34
35			- condvars are used to give blocking behaviour when neccessary. Create
36			a condvar for any "request" or "event" your module might create, C<<
37			->broadcast >> it when the event happens and provide a function that calls
38			C<< ->wait >> for it. See the examples below.
39	root	1.2
40	root	1.1	=head1 DESCRIPTION
41
42	root	1.2	L<AnyEvent> provides an identical interface to multiple event loops. This
43			allows module authors to utilizy an event loop without forcing module
44			users to use the same event loop (as only a single event loop can coexist
45			peacefully at any one time).
46
47			The interface itself is vaguely similar but not identical to the Event
48			module.
49
50			On the first call of any method, the module tries to detect the currently
51			loaded event loop by probing wether any of the following modules is
52			loaded: L<Coro::Event>, L<Event>, L<Glib>, L<Tk>. The first one found is
53			used. If none is found, the module tries to load these modules in the
54			order given. The first one that could be successfully loaded will be
55			used. If still none could be found, it will issue an error.
56
57	root	1.1	=over 4
58
59			=cut
60
61			package AnyEvent;
62
63	root	1.2	no warnings;
64			use strict 'vars';
65	root	1.1	use Carp;
66
67	root	1.5	our $VERSION = 0.3;
68	root	1.2	our $MODEL;
69	root	1.1
70	root	1.2	our $AUTOLOAD;
71			our @ISA;
72	root	1.1
73			my @models = (
74			[Coro => Coro::Event::],
75			[Event => Event::],
76			[Glib => Glib::],
77			[Tk => Tk::],
78			);
79
80	root	1.3	our %method = map +($_ => 1), qw(io timer condvar broadcast wait cancel DESTROY);
81
82	root	1.1	sub AUTOLOAD {
83			$AUTOLOAD =~ s/.*://;
84
85	root	1.3	$method{$AUTOLOAD}
86			or croak "$AUTOLOAD: not a valid method for AnyEvent objects";
87
88	root	1.2	unless ($MODEL) {
89			# check for already loaded models
90			for (@models) {
91			my ($model, $package) = @$_;
92			if (scalar keys %{ *{"$package\::"} }) {
93	root	1.4	eval "require AnyEvent::Impl::$model";
94	root	1.2	last if $MODEL;
95			}
96	root	1.1	}
97
98	root	1.2	unless ($MODEL) {
99			# try to load a model
100
101			for (@models) {
102			my ($model, $package) = @$_;
103	root	1.4	eval "require AnyEvent::Impl::$model";
104	root	1.2	last if $MODEL;
105			}
106
107			$MODEL
108			or die "No event module selected for AnyEvent and autodetect failed. Install any one of these modules: Coro, Event, Glib or Tk.";
109	root	1.1	}
110			}
111
112	root	1.2	@ISA = $MODEL;
113
114			my $class = shift;
115			$class->$AUTOLOAD (@_);
116	root	1.1	}
117
118	root	1.2	=back
119
120			=head1 EXAMPLE
121
122			The following program uses an io watcher to read data from stdin, a timer
123			to display a message once per second, and a condvar to exit the program
124			when the user enters quit:
125
126			use AnyEvent;
127
128			my $cv = AnyEvent->condvar;
129
130			my $io_watcher = AnyEvent->io (fh => \*STDIN, poll => 'r', cb => sub {
131			warn "io event <$_[0]>\n"; # will always output <r>
132			chomp (my $input = <STDIN>); # read a line
133			warn "read: $input\n"; # output what has been read
134			$cv->broadcast if $input =~ /^q/i; # quit program if /^q/i
135			});
136
137			my $time_watcher; # can only be used once
138
139			sub new_timer {
140			$timer = AnyEvent->timer (after => 1, cb => sub {
141			warn "timeout\n"; # print 'timeout' about every second
142			&new_timer; # and restart the time
143			});
144			}
145
146			new_timer; # create first timer
147
148			$cv->wait; # wait until user enters /^q/i
149
150	root	1.5	=head1 REAL-WORLD EXAMPLE
151
152			Consider the L<Net::FCP> module. It features (among others) the following
153			API calls, which are to freenet what HTTP GET requests are to http:
154
155			my $data = $fcp->client_get ($url); # blocks
156
157			my $transaction = $fcp->txn_client_get ($url); # does not block
158			$transaction->cb ( sub { ... } ); # set optional result callback
159			my $data = $transaction->result; # possibly blocks
160
161			The C<client_get> method works like C<LWP::Simple::get>: it requests the
162			given URL and waits till the data has arrived. It is defined to be:
163
164			sub client_get { $_[0]->txn_client_get ($_[1])->result }
165
166			And in fact is automatically generated. This is the blocking API of
167			L<Net::FCP>, and it works as simple as in any other, similar, module.
168
169			More complicated is C<txn_client_get>: It only creates a transaction
170			(completion, result, ...) object and initiates the transaction.
171
172			my $txn = bless { }, Net::FCP::Txn::;
173
174			It also creates a condition variable that is used to signal the completion
175			of the request:
176
177			$txn->{finished} = AnyAvent->condvar;
178
179			It then creates a socket in non-blocking mode.
180
181			socket $txn->{fh}, ...;
182			fcntl $txn->{fh}, F_SETFL, O_NONBLOCK;
183			connect $txn->{fh}, ...
184			and !$!{EWOULDBLOCK}
185			and !$!{EINPROGRESS}
186			and Carp::croak "unable to connect: $!\n";
187
188	root	1.6	Then it creates a write-watcher which gets called whenever an error occurs
189	root	1.5	or the connection succeeds:
190
191			$txn->{w} = AnyEvent->io (fh => $txn->{fh}, poll => 'w', cb => sub { $txn->fh_ready_w });
192
193			And returns this transaction object. The C<fh_ready_w> callback gets
194			called as soon as the event loop detects that the socket is ready for
195			writing.
196
197			The C<fh_ready_w> method makes the socket blocking again, writes the
198			request data and replaces the watcher by a read watcher (waiting for reply
199			data). The actual code is more complicated, but that doesn't matter for
200			this example:
201
202			fcntl $txn->{fh}, F_SETFL, 0;
203			syswrite $txn->{fh}, $txn->{request}
204			or die "connection or write error";
205			$txn->{w} = AnyEvent->io (fh => $txn->{fh}, poll => 'r', cb => sub { $txn->fh_ready_r });
206
207			Again, C<fh_ready_r> waits till all data has arrived, and then stores the
208			result and signals any possible waiters that the request ahs finished:
209
210			sysread $txn->{fh}, $txn->{buf}, length $txn->{$buf};
211
212			if (end-of-file or data complete) {
213			$txn->{result} = $txn->{buf};
214			$txn->{finished}->broadcast;
215	root	1.6	$txb->{cb}->($txn) of $txn->{cb}; # also call callback
216	root	1.5	}
217
218			The C<result> method, finally, just waits for the finished signal (if the
219			request was already finished, it doesn't wait, of course, and returns the
220			data:
221
222			$txn->{finished}->wait;
223	root	1.6	return $txn->{result};
224	root	1.5
225			The actual code goes further and collects all errors (C<die>s, exceptions)
226			that occured during request processing. The C<result> method detects
227			wether an exception as thrown (it is stored inside the $txn object)
228			and just throws the exception, which means connection errors and other
229			problems get reported tot he code that tries to use the result, not in a
230			random callback.
231
232			All of this enables the following usage styles:
233
234			1. Blocking:
235
236			my $data = $fcp->client_get ($url);
237
238			2. Blocking, but parallelizing:
239
240			my @datas = map $_->result,
241			map $fcp->txn_client_get ($_),
242			@urls;
243
244			Both blocking examples work without the module user having to know
245			anything about events.
246
247			3a. Event-based in a main program, using any support Event module:
248
249			use Event;
250
251			$fcp->txn_client_get ($url)->cb (sub {
252			my $txn = shift;
253			my $data = $txn->result;
254			...
255			});
256
257			Event::loop;
258
259			3b. The module user could use AnyEvent, too:
260
261			use AnyEvent;
262
263			my $quit = AnyEvent->condvar;
264
265			$fcp->txn_client_get ($url)->cb (sub {
266			...
267			$quit->broadcast;
268			});
269
270			$quit->wait;
271
272	root	1.2	=head1 SEE ALSO
273
274	root	1.5	Event modules: L<Coro::Event>, L<Coro>, L<Event>, L<Glib::Event>, L<Glib>.
275
276			Implementations: L<AnyEvent::Impl::Coro>, L<AnyEvent::Impl::Event>, L<AnyEvent::Impl::Glib>, L<AnyEvent::Impl::Tk>.
277
278			Nontrivial usage example: L<Net::FCP>.
279	root	1.2
280			=head1
281
282			=cut
283
284			1
285	root	1.1