view release on metacpan or  search on metacpan

local/lib/perl5/IO/Async/Function.pm  view on Meta::CPAN

#  You may distribute under the terms of either the GNU General Public License
#  or the Artistic License (the same terms as Perl itself)
#
#  (C) Paul Evans, 2011-2015 -- leonerd@leonerd.org.uk

package IO::Async::Function;

use strict;
use warnings;

our $VERSION = '0.70';

use base qw( IO::Async::Notifier );
use IO::Async::Timer::Countdown;

use Carp;

=head1 NAME

C<IO::Async::Function> - call a function asynchronously

=head1 SYNOPSIS

 use IO::Async::Function;

 use IO::Async::Loop;
 my $loop = IO::Async::Loop->new;

 my $function = IO::Async::Function->new(
    code => sub {
       my ( $number ) = @_;
       return is_prime( $number );
    },
 );

 $loop->add( $function );

 $function->call(
    args => [ 123454321 ],
 )->on_done( sub {
    my $isprime = shift;
    print "123454321 " . ( $isprime ? "is" : "is not" ) . " a prime number\n";
 })->on_fail( sub {
    print STDERR "Cannot determine if it's prime - $_[0]\n";
 })->get;

=head1 DESCRIPTION

This subclass of L<IO::Async::Notifier> wraps a function body in a collection
of worker processes, to allow it to execute independently of the main process.
The object acts as a proxy to the function, allowing invocations to be made by
passing in arguments, and invoking a continuation in the main process when the
function returns.

The object represents the function code itself, rather than one specific
invocation of it. It can be called multiple times, by the C<call> method.
Multiple outstanding invocations can be called; they will be dispatched in
the order they were queued. If only one worker process is used then results
will be returned in the order they were called. If multiple are used, then
each request will be sent in the order called, but timing differences between
each worker may mean results are returned in a different order.

Since the code block will be called multiple times within the same child
process, it must take care not to modify any of its state that might affect
subsequent calls. Since it executes in a child process, it cannot make any
modifications to the state of the parent program. Therefore, all the data
required to perform its task must be represented in the call arguments, and
all of the result must be represented in the return values.

The Function object is implemented using an L<IO::Async::Routine> with two
L<IO::Async::Channel> objects to pass calls into and results out from it.

The L<IO::Async> framework generally provides mechanisms for multiplexing IO
tasks between different handles, so there aren't many occasions when such an
asynchronous function is necessary. Two cases where this does become useful
are:

=over 4

=item 1.

When a large amount of computationally-intensive work needs to be performed
(for example, the C<is_prime> test in the example in the C<SYNOPSIS>).

=item 2.

When a blocking OS syscall or library-level function needs to be called, and
no nonblocking or asynchronous version is supplied. This is used by
L<IO::Async::Resolver>.

=back

This object is ideal for representing "pure" functions; that is, blocks of
code which have no stateful effect on the process, and whose result depends
only on the arguments passed in. For a more general co-routine ability, see
also L<IO::Async::Routine>.

=cut

=head1 PARAMETERS

The following named parameters may be passed to C<new> or C<configure>:

=head2 code => CODE

The body of the function to execute.

 @result = $code->( @args )

=head2 init_code => CODE

Optional. If defined, this is invoked exactly once in every child process or
thread, after it is created, but before the first invocation of the function
body itself.

 $init_code->()

=head2 model => "fork" | "thread"

Optional. Requests a specific L<IO::Async::Routine> model. If not supplied,
leaves the default choice up to Routine.

=head2 min_workers => INT

=head2 max_workers => INT

The lower and upper bounds of worker processes to try to keep running. The
actual number running at any time will be kept somewhere between these bounds
according to load.

=head2 max_worker_calls => INT

Optional. If provided, stop a worker process after it has processed this
number of calls. (New workers may be started to replace stopped ones, within
the bounds given above).

=head2 idle_timeout => NUM

Optional. If provided, idle worker processes will be shut down after this
amount of time, if there are more than C<min_workers> of them.

=head2 exit_on_die => BOOL

Optional boolean, controls what happens after the C<code> throws an
exception. If missing or false, the worker will continue running to process
more requests. If true, the worker will be shut down. A new worker might be
constructed by the C<call> method to replace it, if necessary.