File Coverage

blib/lib/Net/Async/HTTP/Server.pm

Criterion	Covered	Total	%
statement	59	59	100.0
branch	7	8	87.5
condition	1	2	50.0
subroutine	15	15	100.0
pod	3	3	100.0
total	85	87	97.7

#  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, 2013-2020 -- [e-mail]

package Net::Async::HTTP::Server;

use strict;
use warnings;
use base qw( IO::Async::Listener );
IO::Async::Listener->VERSION( '0.61' );

our $VERSION = '0.13';

use Carp;

use Net::Async::HTTP::Server::Protocol;
use Net::Async::HTTP::Server::Request;

use Metrics::Any 0.05 '$metrics',
   strict      => 1,
   name_prefix => [qw( http server )];

$metrics->make_gauge( requests_in_flight =>
   description => "Count of the number of requests received that have not yet been completed",
   # no labels
);
$metrics->make_counter( requests  =>
   description => "Number of HTTP requests received",
   labels      => [qw( method )],
);
$metrics->make_counter( responses =>
   description => "Number of HTTP responses served",
   labels      => [qw( method code )],
);
$metrics->make_timer( request_duration =>
   description => "Duration of time spent processing requests",
   # no labels
);
$metrics->make_distribution( response_bytes =>
   description => "The size in bytes of responses sent",
   units       => "bytes",
   # no labels
);

=head1 NAME

C - serve HTTP with C

=head1 SYNOPSIS

 use Net::Async::HTTP::Server;
 use IO::Async::Loop;

 use HTTP::Response;

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

 my $httpserver = Net::Async::HTTP::Server->new(
    on_request => sub {
       my $self = shift;
       my ( $req ) = @_;

       my $response = HTTP::Response->new( 200 );
       $response->add_content( "Hello, world!\n" );
       $response->content_type( "text/plain" );
       $response->content_length( length $response->content );

       $req->respond( $response );
    },
 );

 $loop->add( $httpserver );

 $httpserver->listen(
    addr => { family => "inet6", socktype => "stream", port => 8080 },
 )->get

 $loop->run;

=head1 DESCRIPTION

This module allows a program to respond asynchronously to HTTP requests, as
part of a program based on L. An object in this class listens on a
single port and invokes the C callback or subclass method whenever
an HTTP request is received, allowing the program to respond to it.

For accepting HTTP connections via L and L, see also
L.

=head2 Metrics

I

This module reports basic metrics about received requests and sent responses
via L.

=cut

=head1 EVENTS

=head2 on_request $req

Invoked when a new HTTP request is received. It will be passed a
L object.

=cut

=head1 PARAMETERS

The following named parameters may be passed to C or C:

=head2 request_class => STRING

Gives the name of the class that C will construct. This is
provided as an alternative to overriding the C method, for the
case where no other methods need overriding or other behaviour changed.

=cut

=head1 METHODS

As a small subclass of L, this class does not provide many
new methods of its own. The superclass provides useful methods to control the
basic operation of this server.

Specifically, see the L method on how to actually
bind the server to a listening socket to make it accept requests.

=cut

sub _init
{
   my $self = shift;
   my ( $params ) = @_;

   $params->{handle_class} = "Net::Async::HTTP::Server::Protocol";

   $params->{request_class} ||= "Net::Async::HTTP::Server::Request";

   $self->SUPER::_init( $params );
}

sub configure
{
   my $self = shift;
   my %params = @_;

   foreach (qw( on_request request_class )) {
      $self->{$_} = delete $params{$_} if exists $params{$_};
   }

   $self->SUPER::configure( %params );
}

sub _add_to_loop
{
   my $self = shift;

   $self->can_event( "on_request" ) or croak "Expected either a on_request callback or an ->on_request method";

   $self->SUPER::_add_to_loop( @_ );
}

sub on_accept
{
   my $self = shift;
   my ( $conn ) = @_;

   $conn->configure(
      on_closed => sub {
         my $conn = shift;
         $conn->on_closed();

         $conn->remove_from_parent;
      },
   );

   $self->add_child( $conn );

   return $conn;
}

=head2 make_request

   $request = $server->make_request( @args )

Invoked by the protocol stream handler to create a new request object
representing an incoming request. This is provided as a method for subclasses
to overload, if they wish to represent requests with subclasses of the basic
request representation.

=cut

sub make_request
{
   my $self = shift;
   return $self->{request_class}->new( @_ );
}

sub _received_request
{
   my $self = shift;
   my ( $request ) = @_;

   if( $metrics ) {
      $metrics->inc_gauge( requests_in_flight => );

      $metrics->inc_counter( requests => [ method => $request->method ] );
      $self->{request_received_timestamp}{$request} = $self->loop->time;
   }

   $self->invoke_event( on_request => $request );
}

sub _done_request
{
   my $self = shift;
   my ( $request ) = @_;

   if( $metrics ) {
      my $received_timestamp = delete $self->{request_received_timestamp}{$request};

      $metrics->dec_gauge( requests_in_flight => );

      $metrics->inc_counter( responses => [ method => $request->method, code => $request->response_status_code ] );
      $metrics->report_timer( request_duration => $self->loop->time - $received_timestamp );
      $metrics->report_distribution( response_bytes => $request->bytes_written );
   }
}

=head1 TODO

=over 2

=item *

Don't use L objects as underlying implementation

=item *

Consider how to do streaming request inbound

=item *

Lots more testing

=back

=cut

=head1 AUTHOR

Paul Evans

=cut

0x55AA;