Net::Async::FastCGI(3pm) | User Contributed Perl Documentation | Net::Async::FastCGI(3pm) |
"Net::Async::FastCGI" - use FastCGI with IO::Async
As an adapter:
use Net::Async::FastCGI; use IO::Async::Loop; my $loop = IO::Async::Loop->new(); my $fastcgi = Net::Async::FastCGI->new( on_request => sub { my ( $fastcgi, $req ) = @_; # Handle the request here } ); $loop->add( $fastcgi ); $fastcgi->listen( service => 1234, on_resolve_error => sub { die "Cannot resolve - $_[-1]\n" }, on_listen_error => sub { die "Cannot listen - $_[-1]\n" }, ); $loop->run;
As a subclass:
package MyFastCGIResponder; use base qw( Net::Async::FastCGI ); sub on_request { my $self = shift; my ( $req ) = @_; # Handle the request here } ... use IO::Async::Loop; my $loop = IO::Async::Loop->new(); my $fastcgi; $loop->add( $fastcgi = MyFastCGIResponder->new( service => 1234 ) ); $fastcgi->listen( service => 1234, on_resolve_error => sub { die "Cannot resolve - $_[-1]\n" }, on_listen_error => sub { die "Cannot listen - $_[-1]\n" }, ); $loop->run;
This module allows a program to respond asynchronously to FastCGI requests, as part of a program based on IO::Async. An object in this class represents a single FastCGI responder that the webserver is configured to communicate with. It can handle multiple outstanding requests at a time, responding to each as data is provided by the program. Individual outstanding requests that have been started but not yet finished, are represented by instances of Net::Async::FastCGI::Request.
The following events are invoked, either using subclass methods or CODE references in parameters:
Invoked when a new FastCGI request is received. It will be passed a new Net::Async::FastCGI::Request object.
The following named parameters may be passed to "new" or "configure":
Start listening for connections on a socket, creating it first if necessary.
This method may be called in either of the following ways. To listen on an existing socket filehandle:
Or, to create the listening socket or sockets:
This method may also require "on_listen_error" or "on_resolve_error" callbacks for error handling - see IO::Async::Listener for more detail.
The "FCGI_GET_VALUES" FastCGI request can enquire of the responder the maximum number of connections or requests it can support. Because this module puts no fundamental limit on these values, it will return some arbitrary numbers. These are given in package variables:
$Net::Async::FastCGI::MAX_CONNS = 1024; $Net::Async::FastCGI::MAX_REQS = 1024;
These variables are provided in case the containing application wishes to make the library return different values in the request. These values are not actually used by the library, other than to fill in the values in response of "FCGI_GET_VALUES".
When running a local FastCGI responder, the webserver will create a new INET socket connected to the script's STDIN file handle. To use the socket in this case, it should be passed as the "handle" argument.
Paul Evans <leonerd@leonerd.org.uk>
2022-12-10 | perl v5.36.0 |