The Perl Toolchain Summit needs more sponsors. If your company depends on Perl, please support this very important event.

NAME

    Async::Simple::Task::Fork - Forks child process.
    It waits for "data" whic will be passed via "put", and executed "sub" with this "data" as an argument.
    Result of execution will be returned to parent by "get".

SYNOPSIS

    use Async::Simple::Task::Fork;

    my $sub  = sub { sleep 1; return $_[0]{x} + 1 };             # Accepts $data as @_ and returns any type you need

    my $task = Async::Simple::Task::Fork->new( task => &$sub );  # Creates a child process, which waits for data and execute &$sub if data is passed

    my $data = { x => 123 };                                     # Any type you wish: scalar, array, hash

    $task->put( $data );                                         # Put a task data to sub in the child process

    # ...do something useful in parent while our data working ...

    my $result = $task->get; # result = undef because result is not ready yet
    sleep 2; # or do something else....
    my $result = $task->get; # result = 2

    $task->put( $data );                                           # Put another data to task sub and so on,....

Result and data can be of any type and deep which can be translated via Data::Serializer->new( serializer => Storable ) # by default

If your "sub" can return undef you should check $task->has_result, as a mark that result is ready.

DESCRIPTION

Allows to initialize fork process.

After that, executes "sub" for each "data" passed to child process.

METHODS

new

Forks a process

    my $task = Async::Simple::Task::Fork->new( task => &$sub, %other_optional_params );

Params (all param except "task" are optional):

    task         => coderef, function, called for each "data" passed to child process via $task->put( $data );

    timeout      => timeout in seconds between child checkings for new data passed. default 0.01

    kill_on_exit => kill (1) or not (0) subprocess on object destroy (1 by default).

put

Puts data to task.

    $self->put( $data );

get

Tries to read result from task.

Returns undef if it is not ready.

In case, your function can return undef, you shoud check $task->has_answer, as a mark of ready result.

    my $result = $self->get();

SUPPORT AND DOCUMENTATION

    After installing, you can find documentation for this module with the perldoc command.

    perldoc Async::Simple::Task::Fork

    You can also look for information at:

        RT, CPAN's request tracker (report bugs here)
            http://rt.cpan.org/NoAuth/Bugs.html?Dist=Async-Simple-Task-Fork

        AnnoCPAN, Annotated CPAN documentation
            http://annocpan.org/dist/Async-Simple-Task-Fork

        CPAN Ratings
            http://cpanratings.perl.org/d/Async-Simple-Task-Fork

        Search CPAN
            http://search.cpan.org/dist/Async-Simple-Task-Fork/

AUTHOR

ANTONC <antonc@cpan.org>

LICENSE

    This program is free software; you can redistribute it and/or modify it
    under the terms of the the Artistic License (2.0). You may obtain a
    copy of the full license at:

    L<http://www.perlfoundation.org/artistic_license_2_0>

Attributes

task

    task = sub {
        my ( $data ) = @_; # source data for task
        ... your task code ...
        return( $result );
    }

answer

Result of current task

has_answer

has_answer is true, if the task has been finished and result has been ready

timeout

timeout - positive numeric value = seconds between checking for result.

inherited from Async::Simple::Task.

kill_on_exit

Kills process from parent in case of object desctuction

new()

    my $task = Async::Simple::Task::Fork->new( %all_optional_params );

Possible keys for %all_optional_params:

    task         => coderef, function, called for each "data" passed to child process via $task->put( $data );

    timeout      => timeout in seconds between child checkings for new data passed. default 0.01

    kill_on_exit => kill (1) or not (0) subprocess on object destroy (1 by default).

BUILD

internal. Some tricks here:)

    1. Master process called $task->new with fork() inside
    2. After forking done we have two processes:
    2.1. Master gets one side of reader/writer pipes and pid of child
    2.2. Child - another side of pipes and extra logic with everlasting loop

fork_child

Makes child process and returns pid of child process to parent or 0 to child process

get

Reads from task, if something can be readed or returns undef after timeout.

    my $result = $task->get;

Please note! If your function can return an undef value, then you shoud check

    $task->has_result.

put

Writes task to task.

    $task->put( $data );

get_serializer

Internal. Returns an object that must have 2 methods: serialize and deserialize. By default returns Data::Serializer with Storable as backend.

    $self->serializer->serialize( $task_data_ref );

    $result_ref = $self->serializer->deserialize();

DEMOLISH

Destroys object and probably should finish the child process.