[ViewVC] Diff of: cvs/AnyEvent-MP/MP.pm

Comparing AnyEvent-MP/MP.pm (file contents):
Revision 1.7 by root, Sat Aug 1 15:04:30 2009 UTC vs.
Revision 1.104 by root, Fri Nov 6 17:47:20 2009 UTC

…		…
1	=head1 NAME	1	=head1 NAME
2		2
3	AnyEvent::MP - multi-processing/message-passing framework	3	AnyEvent::MP - erlang-style multi-processing/message-passing framework
4		4
5	=head1 SYNOPSIS	5	=head1 SYNOPSIS
6		6
7	use AnyEvent::MP;	7	use AnyEvent::MP;
8		8
9	NODE # returns this node identifier
10	$NODE # contains this node identifier	9	$NODE # contains this node's node ID
		10	NODE # returns this node's node ID
11		11
		12	$SELF # receiving/own port id in rcv callbacks
		13
		14	# initialise the node so it can send/receive messages
		15	configure;
		16
		17	# ports are message destinations
		18
		19	# sending messages
12	snd $port, type => data...;	20	snd $port, type => data...;
		21	snd $port, @msg;
		22	snd @msg_with_first_element_being_a_port;
13		23
14	rcv $port, smartmatch => $cb->($port, @msg);	24	# creating/using ports, the simple way
		25	my $simple_port = port { my @msg = @_ };
15		26
16	# examples:	27	# creating/using ports, tagged message matching
		28	my $port = port;
17	rcv $port2, ping => sub { snd $_[0], "pong"; 0 };	29	rcv $port, ping => sub { snd $_[0], "pong" };
18	rcv $port1, pong => sub { warn "pong received\n" };	30	rcv $port, pong => sub { warn "pong received\n" };
19	snd $port2, ping => $port1;
20		31
21	# more, smarter, matches (_any_ is exported by this module)	32	# create a port on another node
22	rcv $port, [child_died => $pid] => sub { ...	33	my $port = spawn $node, $initfunc, @initdata;
23	rcv $port, [_any_, _any_, 3] => sub { .. $_[2] is 3	34
		35	# destroy a prot again
		36	kil $port; # "normal" kill
		37	kil $port, my_error => "everything is broken"; # error kill
		38
		39	# monitoring
		40	mon $localport, $cb->(@msg) # callback is invoked on death
		41	mon $localport, $otherport # kill otherport on abnormal death
		42	mon $localport, $otherport, @msg # send message on death
		43
		44	# temporarily execute code in port context
		45	peval $port, sub { die "kill the port!" };
		46
		47	# execute callbacks in $SELF port context
		48	my $timer = AE::timer 1, 0, psub {
		49	die "kill the port, delayed";
		50	};
		51
		52	=head1 CURRENT STATUS
		53
		54	bin/aemp - stable.
		55	AnyEvent::MP - stable API, should work.
		56	AnyEvent::MP::Intro - explains most concepts.
		57	AnyEvent::MP::Kernel - mostly stable API.
		58	AnyEvent::MP::Global - stable API.
24		59
25	=head1 DESCRIPTION	60	=head1 DESCRIPTION
26		61
27	This module (-family) implements a simple message passing framework.	62	This module (-family) implements a simple message passing framework.
28		63
29	Despite its simplicity, you can securely message other processes running	64	Despite its simplicity, you can securely message other processes running
30	on the same or other hosts.	65	on the same or other hosts, and you can supervise entities remotely.
31		66
32	At the moment, this module family is severly brokena nd underdocumented,	67	For an introduction to this module family, see the L<AnyEvent::MP::Intro>
33	so do not use. This was uploaded mainly to resreve the CPAN namespace -	68	manual page and the examples under F<eg/>.
34	stay tuned!
35		69
36	=head1 CONCEPTS	70	=head1 CONCEPTS
37		71
38	=over 4	72	=over 4
39		73
40	=item port	74	=item port
41		75
42	A port is something you can send messages to with the C<snd> function, and	76	Not to be confused with a TCP port, a "port" is something you can send
43	you can register C<rcv> handlers with. All C<rcv> handlers will receive	77	messages to (with the C<snd> function).
44	messages they match, messages will not be queued.
45		78
		79	Ports allow you to register C<rcv> handlers that can match all or just
		80	some messages. Messages send to ports will not be queued, regardless of
		81	anything was listening for them or not.
		82
46	=item port id - C<noderef#portname>	83	=item port ID - C<nodeid#portname>
47		84
48	A port id is always the noderef, a hash-mark (C<#>) as separator, followed	85	A port ID is the concatenation of a node ID, a hash-mark (C<#>) as
49	by a port name (a printable string of unspecified format).	86	separator, and a port name (a printable string of unspecified format).
50		87
51	=item node	88	=item node
52		89
53	A node is a single process containing at least one port - the node	90	A node is a single process containing at least one port - the node port,
54	port. You can send messages to node ports to let them create new ports,	91	which enables nodes to manage each other remotely, and to create new
55	among other things.	92	ports.
56		93
57	Initially, nodes are either private (single-process only) or hidden	94	Nodes are either public (have one or more listening ports) or private
58	(connected to a master node only). Only when they epxlicitly "become	95	(no listening ports). Private nodes cannot talk to other private nodes
59	public" can you send them messages from unrelated other nodes.	96	currently.
60		97
61	=item noderef - C<host:port,host:port...>, C<id@noderef>, C<id>	98	=item node ID - C<[A-Z_][a-zA-Z0-9_\-.:]*>
62		99
63	A noderef is a string that either uniquely identifies a given node (for	100	A node ID is a string that uniquely identifies the node within a
64	private and hidden nodes), or contains a recipe on how to reach a given	101	network. Depending on the configuration used, node IDs can look like a
65	node (for public nodes).	102	hostname, a hostname and a port, or a random string. AnyEvent::MP itself
		103	doesn't interpret node IDs in any way.
		104
		105	=item binds - C<ip:port>
		106
		107	Nodes can only talk to each other by creating some kind of connection to
		108	each other. To do this, nodes should listen on one or more local transport
		109	endpoints - binds. Currently, only standard C<ip:port> specifications can
		110	be used, which specify TCP ports to listen on.
		111
		112	=item seed nodes
		113
		114	When a node starts, it knows nothing about the network. To teach the node
		115	about the network it first has to contact some other node within the
		116	network. This node is called a seed.
		117
		118	Apart from the fact that other nodes know them as seed nodes and they have
		119	to have fixed listening addresses, seed nodes are perfectly normal nodes -
		120	any node can function as a seed node for others.
		121
		122	In addition to discovering the network, seed nodes are also used to
		123	maintain the network and to connect nodes that otherwise would have
		124	trouble connecting. They form the backbone of an AnyEvent::MP network.
		125
		126	Seed nodes are expected to be long-running, and at least one seed node
		127	should always be available. They should also be relatively responsive - a
		128	seed node that blocks for long periods will slow down everybody else.
		129
		130	=item seeds - C<host:port>
		131
		132	Seeds are transport endpoint(s) (usually a hostname/IP address and a
		133	TCP port) of nodes that should be used as seed nodes.
		134
		135	The nodes listening on those endpoints are expected to be long-running,
		136	and at least one of those should always be available. When nodes run out
		137	of connections (e.g. due to a network error), they try to re-establish
		138	connections to some seednodes again to join the network.
66		139
67	=back	140	=back
68		141
69	=head1 VARIABLES/FUNCTIONS	142	=head1 VARIABLES/FUNCTIONS
70		143
72		145
73	=cut	146	=cut
74		147
75	package AnyEvent::MP;	148	package AnyEvent::MP;
76		149
77	use AnyEvent::MP::Util ();
78	use AnyEvent::MP::Node;	150	use AnyEvent::MP::Kernel;
79	use AnyEvent::MP::Transport;
80		151
81	use utf8;
82	use common::sense;	152	use common::sense;
83		153
84	use Carp ();	154	use Carp ();
85		155
86	use AE ();	156	use AE ();
87		157
88	use base "Exporter";	158	use base "Exporter";
89		159
90	our $VERSION = '0.01';	160	our $VERSION = 1.23;
91	our @EXPORT = qw(NODE $NODE $PORT snd rcv _any_);
92		161
93	our $DEFAULT_SECRET;	162	our @EXPORT = qw(
94	our $DEFAULT_PORT = "4040";	163	NODE $NODE *SELF node_of after
		164	configure
		165	snd rcv mon mon_guard kil psub peval spawn cal
		166	port
		167	);
95		168
96	our $CONNECT_INTERVAL = 5; # new connect every 5s, at least	169	our $SELF;
97	our $CONNECT_TIMEOUT = 30; # includes handshake
98		170
99	sub default_secret {	171	sub _self_die() {
100	unless (defined $DEFAULT_SECRET) {	172	my $msg = $@;
101	if (open my $fh, "<$ENV{HOME}/.aemp-secret") {	173	$msg =~ s/\n+$// unless ref $msg;
102	sysread $fh, $DEFAULT_SECRET, -s $fh;	174	kil $SELF, die => $msg;
103	} else {
104	$DEFAULT_SECRET = AnyEvent::MP::Util::nonce 32;
105	}
106	}
107
108	$DEFAULT_SECRET
109	}	175	}
110		176
111	=item NODE / $NODE	177	=item $thisnode = NODE / $NODE
112		178
113	The C<NODE ()> function and the C<$NODE> variable contain the noderef of	179	The C<NODE> function returns, and the C<$NODE> variable contains, the node
114	the local node. The value is initialised by a call to C<become_public> or	180	ID of the node running in the current process. This value is initialised by
115	C<become_slave>, after which all local port identifiers become invalid.	181	a call to C<configure>.
116		182
117	=cut	183	=item $nodeid = node_of $port
118		184
119	our $UNIQ = sprintf "%x.%x", $$, time; # per-process/node unique cookie	185	Extracts and returns the node ID from a port ID or a node ID.
120	our $ID = "a0";
121	our $PUBLIC = 0;
122	our $NODE;
123	our $PORT;
124		186
125	our %NODE; # node id to transport mapping, or "undef", for local node	187	=item configure $profile, key => value...
126	our %PORT; # local ports
127	our %LISTENER; # local transports
128		188
129	sub NODE() { $NODE }	189	=item configure key => value...
130		190
131	{	191	Before a node can talk to other nodes on the network (i.e. enter
132	use POSIX ();	192	"distributed mode") it has to configure itself - the minimum a node needs
133	my $nodename = (POSIX::uname)[1];	193	to know is its own name, and optionally it should know the addresses of
134	$NODE = "$$\@$nodename";	194	some other nodes in the network to discover other nodes.
135	}
136		195
137	sub _ANY_() { 1 }	196	This function configures a node - it must be called exactly once (or
138	sub _any_() { \&_ANY_ }	197	never) before calling other AnyEvent::MP functions.
139		198
140	sub add_node {	199	=over 4
141	my ($noderef) = @_;
142		200
143	return $NODE{$noderef}	201	=item step 1, gathering configuration from profiles
144	if exists $NODE{$noderef};
145		202
146	for (split /,/, $noderef) {	203	The function first looks up a profile in the aemp configuration (see the
147	return $NODE{$noderef} = $NODE{$_}	204	L<aemp> commandline utility). The profile name can be specified via the
148	if exists $NODE{$_};	205	named C<profile> parameter or can simply be the first parameter). If it is
149	}	206	missing, then the nodename (F<uname -n>) will be used as profile name.
150		207
151	# for indirect sends, use a different class	208	The profile data is then gathered as follows:
152	my $node = new AnyEvent::MP::Node::Direct $noderef;
153		209
154	$NODE{$_} = $node	210	First, all remaining key => value pairs (all of which are conveniently
155	for $noderef, split /,/, $noderef;	211	undocumented at the moment) will be interpreted as configuration
		212	data. Then they will be overwritten by any values specified in the global
		213	default configuration (see the F<aemp> utility), then the chain of
		214	profiles chosen by the profile name (and any C<parent> attributes).
156		215
157	$node	216	That means that the values specified in the profile have highest priority
158	}	217	and the values specified directly via C<configure> have lowest priority,
		218	and can only be used to specify defaults.
159		219
		220	If the profile specifies a node ID, then this will become the node ID of
		221	this process. If not, then the profile name will be used as node ID. The
		222	special node ID of C<anon/> will be replaced by a random node ID.
		223
		224	=item step 2, bind listener sockets
		225
		226	The next step is to look up the binds in the profile, followed by binding
		227	aemp protocol listeners on all binds specified (it is possible and valid
		228	to have no binds, meaning that the node cannot be contacted form the
		229	outside. This means the node cannot talk to other nodes that also have no
		230	binds, but it can still talk to all "normal" nodes).
		231
		232	If the profile does not specify a binds list, then a default of C<*> is
		233	used, meaning the node will bind on a dynamically-assigned port on every
		234	local IP address it finds.
		235
		236	=item step 3, connect to seed nodes
		237
		238	As the last step, the seeds list from the profile is passed to the
		239	L<AnyEvent::MP::Global> module, which will then use it to keep
		240	connectivity with at least one node at any point in time.
		241
		242	=back
		243
		244	Example: become a distributed node using the local node name as profile.
		245	This should be the most common form of invocation for "daemon"-type nodes.
		246
		247	configure
		248
		249	Example: become an anonymous node. This form is often used for commandline
		250	clients.
		251
		252	configure nodeid => "anon/";
		253
		254	Example: configure a node using a profile called seed, which si suitable
		255	for a seed node as it binds on all local addresses on a fixed port (4040,
		256	customary for aemp).
		257
		258	# use the aemp commandline utility
		259	# aemp profile seed nodeid anon/ binds '*:4040'
		260
		261	# then use it
		262	configure profile => "seed";
		263
		264	# or simply use aemp from the shell again:
		265	# aemp run profile seed
		266
		267	# or provide a nicer-to-remember nodeid
		268	# aemp run profile seed nodeid "$(hostname)"
		269
		270	=item $SELF
		271
		272	Contains the current port id while executing C<rcv> callbacks or C<psub>
		273	blocks.
		274
		275	=item *SELF, SELF, %SELF, @SELF...
		276
		277	Due to some quirks in how perl exports variables, it is impossible to
		278	just export C<$SELF>, all the symbols named C<SELF> are exported by this
		279	module, but only C<$SELF> is currently used.
		280
160	=item snd $portid, type => @data	281	=item snd $port, type => @data
161		282
162	=item snd $portid, @msg	283	=item snd $port, @msg
163		284
164	Send the given message to the given port ID, which can identify either a	285	Send the given message to the given port, which can identify either a
165	local or a remote port.	286	local or a remote port, and must be a port ID.
166		287
167	While the message can be about anything, it is highly recommended to use	288	While the message can be almost anything, it is highly recommended to
168	a constant string as first element.	289	use a string as first element (a port ID, or some word that indicates a
		290	request type etc.) and to consist if only simple perl values (scalars,
		291	arrays, hashes) - if you think you need to pass an object, think again.
169		292
170	The message data effectively becomes read-only after a call to this	293	The message data logically becomes read-only after a call to this
171	function: modifying any argument is not allowed and can cause many	294	function: modifying any argument (or values referenced by them) is
172	problems.	295	forbidden, as there can be considerable time between the call to C<snd>
		296	and the time the message is actually being serialised - in fact, it might
		297	never be copied as within the same process it is simply handed to the
		298	receiving port.
173		299
174	The type of data you can transfer depends on the transport protocol: when	300	The type of data you can transfer depends on the transport protocol: when
175	JSON is used, then only strings, numbers and arrays and hashes consisting	301	JSON is used, then only strings, numbers and arrays and hashes consisting
176	of those are allowed (no objects). When Storable is used, then anything	302	of those are allowed (no objects). When Storable is used, then anything
177	that Storable can serialise and deserialise is allowed, and for the local	303	that Storable can serialise and deserialise is allowed, and for the local
178	node, anything can be passed.	304	node, anything can be passed. Best rely only on the common denominator of
		305	these.
179		306
180	=cut	307	=item $local_port = port
181		308
182	sub snd(@) {	309	Create a new local port object and returns its port ID. Initially it has
183	my ($noderef, $port) = split /#/, shift, 2;	310	no callbacks set and will throw an error when it receives messages.
184		311
185	add_node $noderef	312	=item $local_port = port { my @msg = @_ }
186	unless exists $NODE{$noderef};
187		313
188	$NODE{$noderef}->send (["$port", [@_]]);	314	Creates a new local port, and returns its ID. Semantically the same as
189	}	315	creating a port and calling C<rcv $port, $callback> on it.
190		316
		317	The block will be called for every message received on the port, with the
		318	global variable C<$SELF> set to the port ID. Runtime errors will cause the
		319	port to be C<kil>ed. The message will be passed as-is, no extra argument
		320	(i.e. no port ID) will be passed to the callback.
		321
		322	If you want to stop/destroy the port, simply C<kil> it:
		323
		324	my $port = port {
		325	my @msg = @_;
		326	...
		327	kil $SELF;
		328	};
		329
		330	=cut
		331
		332	sub rcv($@);
		333
		334	sub _kilme {
		335	die "received message on port without callback";
		336	}
		337
		338	sub port(;&) {
		339	my $id = "$UNIQ." . $ID++;
		340	my $port = "$NODE#$id";
		341
		342	rcv $port, shift \|\| \&_kilme;
		343
		344	$port
		345	}
		346
191	=item rcv $portid, type => $callback->(@msg)	347	=item rcv $local_port, $callback->(@msg)
192		348
193	=item rcv $portid, $smartmatch => $callback->(@msg)	349	Replaces the default callback on the specified port. There is no way to
		350	remove the default callback: use C<sub { }> to disable it, or better
		351	C<kil> the port when it is no longer needed.
194		352
195	=item rcv $portid, [$smartmatch...] => $callback->(@msg)	353	The global C<$SELF> (exported by this module) contains C<$port> while
		354	executing the callback. Runtime errors during callback execution will
		355	result in the port being C<kil>ed.
196		356
197	Register a callback on the port identified by C<$portid>, which I<must> be	357	The default callback received all messages not matched by a more specific
198	a local port.	358	C<tag> match.
199		359
200	The callback has to return a true value when its work is done, after	360	=item rcv $local_port, tag => $callback->(@msg_without_tag), ...
201	which is will be removed, or a false value in which case it will stay
202	registered.
203		361
204	If the match is an array reference, then it will be matched against the	362	Register (or replace) callbacks to be called on messages starting with the
205	first elements of the message, otherwise only the first element is being	363	given tag on the given port (and return the port), or unregister it (when
206	matched.	364	C<$callback> is C<$undef> or missing). There can only be one callback
		365	registered for each tag.
207		366
208	Any element in the match that is specified as C<_any_> (a function	367	The original message will be passed to the callback, after the first
209	exported by this module) matches any single element of the message.	368	element (the tag) has been removed. The callback will use the same
		369	environment as the default callback (see above).
210		370
211	While not required, it is highly recommended that the first matching	371	Example: create a port and bind receivers on it in one go.
212	element is a string identifying the message. The one-string-only match is	372
213	also the most efficient match (by far).	373	my $port = rcv port,
		374	msg1 => sub { ... },
		375	msg2 => sub { ... },
		376	;
		377
		378	Example: create a port, bind receivers and send it in a message elsewhere
		379	in one go:
		380
		381	snd $otherport, reply =>
		382	rcv port,
		383	msg1 => sub { ... },
		384	...
		385	;
		386
		387	Example: temporarily register a rcv callback for a tag matching some port
		388	(e.g. for an rpc reply) and unregister it after a message was received.
		389
		390	rcv $port, $otherport => sub {
		391	my @reply = @_;
		392
		393	rcv $SELF, $otherport;
		394	};
214		395
215	=cut	396	=cut
216		397
217	sub rcv($@) {	398	sub rcv($@) {
218	my ($port, $match, $cb) = @_;	399	my $port = shift;
219
220	my $port = $PORT{$port}
221	or do {
222	my ($noderef, $lport) = split /#/, $port;	400	my ($nodeid, $portid) = split /#/, $port, 2;
223	"AnyEvent::MP::Node::Self" eq ref $NODE{$noderef}	401
		402	$NODE{$nodeid} == $NODE{""}
224	or Carp::croak "$port: can only rcv on local ports";	403	or Carp::croak "$port: rcv can only be called on local ports, caught";
225		404
226	$PORT{$lport}	405	while (@_) {
227	or Carp::croak "$port: port does not exist";	406	if (ref $_[0]) {
228		407	if (my $self = $PORT_DATA{$portid}) {
229	$PORT{$port} = $PORT{$lport} # also return	408	"AnyEvent::MP::Port" eq ref $self
230	};	409	or Carp::croak "$port: rcv can only be called on message matching ports, caught";
231		410
232	if (!ref $match) {	411	$self->[0] = shift;
233	push @{ $port->{rc0}{$match} }, [$cb];
234	} elsif (("ARRAY" eq ref $match && !ref $match->[0])) {
235	my ($type, @match) = @$match;
236	@match
237	? push @{ $port->{rcv}{$match->[0]} }, [$cb, \@match]
238	: push @{ $port->{rc0}{$match->[0]} }, [$cb];
239	} else {	412	} else {
240	push @{ $port->{any} }, [$cb, $match];	413	my $cb = shift;
241	}	414	$PORT{$portid} = sub {
242	}	415	local $SELF = $port;
243		416	eval { &$cb }; _self_die if $@;
244	sub _inject {
245	my ($port, $msg) = @{+shift};
246
247	$port = $PORT{$port}
248	or return;
249
250	@_ = @$msg;
251
252	for (@{ $port->{rc0}{$msg->[0]} }) {
253	$_ && &{$_->[0]}
254	&& undef $_;
255	}
256
257	for (@{ $port->{rcv}{$msg->[0]} }) {
258	$_ && [@_[1..$#{$_->[1]}]] ~~ $_->[1]
259	&& &{$_->[0]}
260	&& undef $_;
261	}
262
263	for (@{ $port->{any} }) {
264	$_ && [@_[0..$#{$_->[1]}]] ~~ $_->[1]
265	&& &{$_->[0]}
266	&& undef $_;
267	}
268	}
269
270	sub normalise_noderef($) {
271	my ($noderef) = @_;
272
273	my $cv = AE::cv;
274	my @res;
275
276	$cv->begin (sub {
277	my %seen;
278	my @refs;
279	for (sort { $a->[0] <=> $b->[0] } @res) {
280	push @refs, $_->[1] unless $seen{$_->[1]}++
281	}
282	shift->send (join ",", @refs);
283	});
284
285	$noderef = $DEFAULT_PORT unless length $noderef;
286
287	my $idx;
288	for my $t (split /,/, $noderef) {
289	my $pri = ++$idx;
290
291	#TODO: this should be outside normalise_noderef and in become_public
292	if ($t =~ /^\d*$/) {
293	my $nodename = (POSIX::uname)[1];
294
295	$cv->begin;
296	AnyEvent::Socket::resolve_sockaddr $nodename, $t \|\| "aemp=$DEFAULT_PORT", "tcp", 0, undef, sub {
297	for (@_) {
298	my ($service, $host) = AnyEvent::Socket::unpack_sockaddr $_->[3];
299	push @res, [
300	$pri += 1e-5,
301	AnyEvent::Socket::format_hostport AnyEvent::Socket::format_address $host, $service
302	];
303	}	417	};
304	$cv->end;	418	}
		419	} elsif (defined $_[0]) {
		420	my $self = $PORT_DATA{$portid} \|\|= do {
		421	my $self = bless [$PORT{$portid} \|\| sub { }, { }, $port], "AnyEvent::MP::Port";
		422
		423	$PORT{$portid} = sub {
		424	local $SELF = $port;
		425
		426	if (my $cb = $self->[1]{$_[0]}) {
		427	shift;
		428	eval { &$cb }; _self_die if $@;
		429	} else {
		430	&{ $self->[0] };
		431	}
		432	};
		433
		434	$self
305	};	435	};
306		436
307	# my (undef, undef, undef, undef, @ipv4) = gethostbyname $nodename;	437	"AnyEvent::MP::Port" eq ref $self
308	#	438	or Carp::croak "$port: rcv can only be called on message matching ports, caught";
309	# for (@ipv4) {	439
310	# push @res, [	440	my ($tag, $cb) = splice @_, 0, 2;
311	# $pri,	441
312	# AnyEvent::Socket::format_hostport AnyEvent::Socket::format_address $_, $t \|\| $DEFAULT_PORT,	442	if (defined $cb) {
313	# ];	443	$self->[1]{$tag} = $cb;
314	# }
315	} else {	444	} else {
316	my ($host, $port) = AnyEvent::Socket::parse_hostport $t, "aemp=$DEFAULT_PORT"	445	delete $self->[1]{$tag};
317	or Carp::croak "$t: unparsable transport descriptor";
318
319	$cv->begin;
320	AnyEvent::Socket::resolve_sockaddr $host, $port, "tcp", 0, undef, sub {
321	for (@_) {
322	my ($service, $host) = AnyEvent::Socket::unpack_sockaddr $_->[3];
323	push @res, [
324	$pri += 1e-5,
325	AnyEvent::Socket::format_hostport AnyEvent::Socket::format_address $host, $service
326	];
327	}
328	$cv->end;
329	}	446	}
330	}	447	}
331	}	448	}
332		449
333	$cv->end;	450	$port
334
335	$cv
336	}	451	}
337		452
338	sub become_public {	453	=item peval $port, $coderef[, @args]
339	return if $PUBLIC;
340		454
341	my $noderef = join ",", ref $_[0] ? @{+shift} : shift;	455	Evaluates the given C<$codref> within the contetx of C<$port>, that is,
342	my @args = @_;	456	when the code throews an exception the C<$port> will be killed.
343		457
344	$NODE = (normalise_noderef $noderef)->recv;	458	Any remaining args will be passed to the callback. Any return values will
		459	be returned to the caller.
345		460
346	for my $t (split /,/, $NODE) {	461	This is useful when you temporarily want to execute code in the context of
347	$NODE{$t} = $NODE{""};	462	a port.
348		463
349	my ($host, $port) = AnyEvent::Socket::parse_hostport $t;	464	Example: create a port and run some initialisation code in it's context.
350		465
351	$LISTENER{$t} = AnyEvent::MP::Transport::mp_server $host, $port,	466	my $port = port { ... };
352	@args,
353	on_error => sub {
354	die "on_error<@_>\n";#d#
355	},
356	on_connect => sub {
357	my ($tp) = @_;
358		467
359	$NODE{$tp->{remote_id}} = $_[0];	468	peval $port, sub {
360	},	469	init
361	sub {	470	or die "unable to init";
362	my ($tp) = @_;	471	};
363		472
364	$NODE{"$tp->{peerhost}:$tp->{peerport}"} = $tp;	473	=cut
365	},	474
366	;	475	sub peval($$) {
		476	local $SELF = shift;
		477	my $cb = shift;
		478
		479	if (wantarray) {
		480	my @res = eval { &$cb };
		481	_self_die if $@;
		482	@res
		483	} else {
		484	my $res = eval { &$cb };
		485	_self_die if $@;
		486	$res
367	}	487	}
		488	}
368		489
369	$PUBLIC = 1;	490	=item $closure = psub { BLOCK }
		491
		492	Remembers C<$SELF> and creates a closure out of the BLOCK. When the
		493	closure is executed, sets up the environment in the same way as in C<rcv>
		494	callbacks, i.e. runtime errors will cause the port to get C<kil>ed.
		495
		496	The effect is basically as if it returned C<< sub { peval $SELF, sub {
		497	BLOCK } } >>.
		498
		499	This is useful when you register callbacks from C<rcv> callbacks:
		500
		501	rcv delayed_reply => sub {
		502	my ($delay, @reply) = @_;
		503	my $timer = AE::timer $delay, 0, psub {
		504	snd @reply, $SELF;
		505	};
		506	};
		507
		508	=cut
		509
		510	sub psub(&) {
		511	my $cb = shift;
		512
		513	my $port = $SELF
		514	or Carp::croak "psub can only be called from within rcv or psub callbacks, not";
		515
		516	sub {
		517	local $SELF = $port;
		518
		519	if (wantarray) {
		520	my @res = eval { &$cb };
		521	_self_die if $@;
		522	@res
		523	} else {
		524	my $res = eval { &$cb };
		525	_self_die if $@;
		526	$res
		527	}
		528	}
		529	}
		530
		531	=item $guard = mon $port, $cb->(@reason) # call $cb when $port dies
		532
		533	=item $guard = mon $port, $rcvport # kill $rcvport when $port dies
		534
		535	=item $guard = mon $port # kill $SELF when $port dies
		536
		537	=item $guard = mon $port, $rcvport, @msg # send a message when $port dies
		538
		539	Monitor the given port and do something when the port is killed or
		540	messages to it were lost, and optionally return a guard that can be used
		541	to stop monitoring again.
		542
		543	In the first form (callback), the callback is simply called with any
		544	number of C<@reason> elements (no @reason means that the port was deleted
		545	"normally"). Note also that I<< the callback B<must> never die >>, so use
		546	C<eval> if unsure.
		547
		548	In the second form (another port given), the other port (C<$rcvport>)
		549	will be C<kil>'ed with C<@reason>, if a @reason was specified, i.e. on
		550	"normal" kils nothing happens, while under all other conditions, the other
		551	port is killed with the same reason.
		552
		553	The third form (kill self) is the same as the second form, except that
		554	C<$rvport> defaults to C<$SELF>.
		555
		556	In the last form (message), a message of the form C<@msg, @reason> will be
		557	C<snd>.
		558
		559	Monitoring-actions are one-shot: once messages are lost (and a monitoring
		560	alert was raised), they are removed and will not trigger again.
		561
		562	As a rule of thumb, monitoring requests should always monitor a port from
		563	a local port (or callback). The reason is that kill messages might get
		564	lost, just like any other message. Another less obvious reason is that
		565	even monitoring requests can get lost (for example, when the connection
		566	to the other node goes down permanently). When monitoring a port locally
		567	these problems do not exist.
		568
		569	C<mon> effectively guarantees that, in the absence of hardware failures,
		570	after starting the monitor, either all messages sent to the port will
		571	arrive, or the monitoring action will be invoked after possible message
		572	loss has been detected. No messages will be lost "in between" (after
		573	the first lost message no further messages will be received by the
		574	port). After the monitoring action was invoked, further messages might get
		575	delivered again.
		576
		577	Inter-host-connection timeouts and monitoring depend on the transport
		578	used. The only transport currently implemented is TCP, and AnyEvent::MP
		579	relies on TCP to detect node-downs (this can take 10-15 minutes on a
		580	non-idle connection, and usually around two hours for idle connections).
		581
		582	This means that monitoring is good for program errors and cleaning up
		583	stuff eventually, but they are no replacement for a timeout when you need
		584	to ensure some maximum latency.
		585
		586	Example: call a given callback when C<$port> is killed.
		587
		588	mon $port, sub { warn "port died because of <@_>\n" };
		589
		590	Example: kill ourselves when C<$port> is killed abnormally.
		591
		592	mon $port;
		593
		594	Example: send us a restart message when another C<$port> is killed.
		595
		596	mon $port, $self => "restart";
		597
		598	=cut
		599
		600	sub mon {
		601	my ($nodeid, $port) = split /#/, shift, 2;
		602
		603	my $node = $NODE{$nodeid} \|\| add_node $nodeid;
		604
		605	my $cb = @_ ? shift : $SELF \|\| Carp::croak 'mon: called with one argument only, but $SELF not set,';
		606
		607	unless (ref $cb) {
		608	if (@_) {
		609	# send a kill info message
		610	my (@msg) = ($cb, @_);
		611	$cb = sub { snd @msg, @_ };
		612	} else {
		613	# simply kill other port
		614	my $port = $cb;
		615	$cb = sub { kil $port, @_ if @_ };
		616	}
		617	}
		618
		619	$node->monitor ($port, $cb);
		620
		621	defined wantarray
		622	and ($cb += 0, AnyEvent::Util::guard { $node->unmonitor ($port, $cb) })
		623	}
		624
		625	=item $guard = mon_guard $port, $ref, $ref...
		626
		627	Monitors the given C<$port> and keeps the passed references. When the port
		628	is killed, the references will be freed.
		629
		630	Optionally returns a guard that will stop the monitoring.
		631
		632	This function is useful when you create e.g. timers or other watchers and
		633	want to free them when the port gets killed (note the use of C<psub>):
		634
		635	$port->rcv (start => sub {
		636	my $timer; $timer = mon_guard $port, AE::timer 1, 1, psub {
		637	undef $timer if 0.9 < rand;
		638	});
		639	});
		640
		641	=cut
		642
		643	sub mon_guard {
		644	my ($port, @refs) = @_;
		645
		646	#TODO: mon-less form?
		647
		648	mon $port, sub { 0 && @refs }
		649	}
		650
		651	=item kil $port[, @reason]
		652
		653	Kill the specified port with the given C<@reason>.
		654
		655	If no C<@reason> is specified, then the port is killed "normally" (ports
		656	monitoring other ports will not necessarily die because a port dies
		657	"normally").
		658
		659	Otherwise, linked ports get killed with the same reason (second form of
		660	C<mon>, see above).
		661
		662	Runtime errors while evaluating C<rcv> callbacks or inside C<psub> blocks
		663	will be reported as reason C<< die => $@ >>.
		664
		665	Transport/communication errors are reported as C<< transport_error =>
		666	$message >>.
		667
		668	=cut
		669
		670	=item $port = spawn $node, $initfunc[, @initdata]
		671
		672	Creates a port on the node C<$node> (which can also be a port ID, in which
		673	case it's the node where that port resides).
		674
		675	The port ID of the newly created port is returned immediately, and it is
		676	possible to immediately start sending messages or to monitor the port.
		677
		678	After the port has been created, the init function is called on the remote
		679	node, in the same context as a C<rcv> callback. This function must be a
		680	fully-qualified function name (e.g. C<MyApp::Chat::Server::init>). To
		681	specify a function in the main program, use C<::name>.
		682
		683	If the function doesn't exist, then the node tries to C<require>
		684	the package, then the package above the package and so on (e.g.
		685	C<MyApp::Chat::Server>, C<MyApp::Chat>, C<MyApp>) until the function
		686	exists or it runs out of package names.
		687
		688	The init function is then called with the newly-created port as context
		689	object (C<$SELF>) and the C<@initdata> values as arguments. It I<must>
		690	call one of the C<rcv> functions to set callbacks on C<$SELF>, otherwise
		691	the port might not get created.
		692
		693	A common idiom is to pass a local port, immediately monitor the spawned
		694	port, and in the remote init function, immediately monitor the passed
		695	local port. This two-way monitoring ensures that both ports get cleaned up
		696	when there is a problem.
		697
		698	C<spawn> guarantees that the C<$initfunc> has no visible effects on the
		699	caller before C<spawn> returns (by delaying invocation when spawn is
		700	called for the local node).
		701
		702	Example: spawn a chat server port on C<$othernode>.
		703
		704	# this node, executed from within a port context:
		705	my $server = spawn $othernode, "MyApp::Chat::Server::connect", $SELF;
		706	mon $server;
		707
		708	# init function on C<$othernode>
		709	sub connect {
		710	my ($srcport) = @_;
		711
		712	mon $srcport;
		713
		714	rcv $SELF, sub {
		715	...
		716	};
		717	}
		718
		719	=cut
		720
		721	sub _spawn {
		722	my $port = shift;
		723	my $init = shift;
		724
		725	# rcv will create the actual port
		726	local $SELF = "$NODE#$port";
		727	eval {
		728	&{ load_func $init }
		729	};
		730	_self_die if $@;
		731	}
		732
		733	sub spawn(@) {
		734	my ($nodeid, undef) = split /#/, shift, 2;
		735
		736	my $id = "$RUNIQ." . $ID++;
		737
		738	$_[0] =~ /::/
		739	or Carp::croak "spawn init function must be a fully-qualified name, caught";
		740
		741	snd_to_func $nodeid, "AnyEvent::MP::_spawn" => $id, @_;
		742
		743	"$nodeid#$id"
		744	}
		745
		746	=item after $timeout, @msg
		747
		748	=item after $timeout, $callback
		749
		750	Either sends the given message, or call the given callback, after the
		751	specified number of seconds.
		752
		753	This is simply a utility function that comes in handy at times - the
		754	AnyEvent::MP author is not convinced of the wisdom of having it, though,
		755	so it may go away in the future.
		756
		757	=cut
		758
		759	sub after($@) {
		760	my ($timeout, @action) = @_;
		761
		762	my $t; $t = AE::timer $timeout, 0, sub {
		763	undef $t;
		764	ref $action[0]
		765	? $action[0]()
		766	: snd @action;
		767	};
		768	}
		769
		770	=item cal $port, @msg, $callback[, $timeout]
		771
		772	A simple form of RPC - sends a message to the given C<$port> with the
		773	given contents (C<@msg>), but adds a reply port to the message.
		774
		775	The reply port is created temporarily just for the purpose of receiving
		776	the reply, and will be C<kil>ed when no longer needed.
		777
		778	A reply message sent to the port is passed to the C<$callback> as-is.
		779
		780	If an optional time-out (in seconds) is given and it is not C<undef>,
		781	then the callback will be called without any arguments after the time-out
		782	elapsed and the port is C<kil>ed.
		783
		784	If no time-out is given (or it is C<undef>), then the local port will
		785	monitor the remote port instead, so it eventually gets cleaned-up.
		786
		787	Currently this function returns the temporary port, but this "feature"
		788	might go in future versions unless you can make a convincing case that
		789	this is indeed useful for something.
		790
		791	=cut
		792
		793	sub cal(@) {
		794	my $timeout = ref $_[-1] ? undef : pop;
		795	my $cb = pop;
		796
		797	my $port = port {
		798	undef $timeout;
		799	kil $SELF;
		800	&$cb;
		801	};
		802
		803	if (defined $timeout) {
		804	$timeout = AE::timer $timeout, 0, sub {
		805	undef $timeout;
		806	kil $port;
		807	$cb->();
		808	};
		809	} else {
		810	mon $_[0], sub {
		811	kil $port;
		812	$cb->();
		813	};
		814	}
		815
		816	push @_, $port;
		817	&snd;
		818
		819	$port
370	}	820	}
371		821
372	=back	822	=back
373		823
374	=head1 NODE MESSAGES	824	=head1 AnyEvent::MP vs. Distributed Erlang
375		825
376	Nodes understand the following messages sent to them. Many of them take	826	AnyEvent::MP got lots of its ideas from distributed Erlang (Erlang node
377	arguments called C<@reply>, which will simply be used to compose a reply	827	== aemp node, Erlang process == aemp port), so many of the documents and
378	message - C<$reply[0]> is the port to reply to, C<$reply[1]> the type and	828	programming techniques employed by Erlang apply to AnyEvent::MP. Here is a
379	the remaining arguments are simply the message data.	829	sample:
		830
		831	http://www.erlang.se/doc/programming_rules.shtml
		832	http://erlang.org/doc/getting_started/part_frame.html # chapters 3 and 4
		833	http://erlang.org/download/erlang-book-part1.pdf # chapters 5 and 6
		834	http://erlang.org/download/armstrong_thesis_2003.pdf # chapters 4 and 5
		835
		836	Despite the similarities, there are also some important differences:
380		837
381	=over 4	838	=over 4
382		839
383	=cut	840	=item * Node IDs are arbitrary strings in AEMP.
384		841
385	#############################################################################	842	Erlang relies on special naming and DNS to work everywhere in the same
386	# self node code	843	way. AEMP relies on each node somehow knowing its own address(es) (e.g. by
		844	configuration or DNS), and possibly the addresses of some seed nodes, but
		845	will otherwise discover other nodes (and their IDs) itself.
387		846
388	sub _new_port($) {	847	=item * Erlang has a "remote ports are like local ports" philosophy, AEMP
389	my ($name) = @_;	848	uses "local ports are like remote ports".
390		849
391	my ($noderef, $portname) = split /#/, $name;	850	The failure modes for local ports are quite different (runtime errors
		851	only) then for remote ports - when a local port dies, you I<know> it dies,
		852	when a connection to another node dies, you know nothing about the other
		853	port.
392		854
393	$PORT{$name} =	855	Erlang pretends remote ports are as reliable as local ports, even when
394	$PORT{$portname} = {	856	they are not.
395	names => [$name, $portname],
396	};
397	}
398		857
399	$NODE{""} = new AnyEvent::MP::Node::Self noderef => $NODE;	858	AEMP encourages a "treat remote ports differently" philosophy, with local
400	_new_port "";	859	ports being the special case/exception, where transport errors cannot
		860	occur.
401		861
402	=item devnull => ...	862	=item * Erlang uses processes and a mailbox, AEMP does not queue.
403		863
404	Generic data sink/CPU heat conversion.	864	Erlang uses processes that selectively receive messages, and therefore
		865	needs a queue. AEMP is event based, queuing messages would serve no
		866	useful purpose. For the same reason the pattern-matching abilities of
		867	AnyEvent::MP are more limited, as there is little need to be able to
		868	filter messages without dequeuing them.
405		869
406	=cut	870	(But see L<Coro::MP> for a more Erlang-like process model on top of AEMP).
407		871
408	rcv "", devnull => sub { () };	872	=item * Erlang sends are synchronous, AEMP sends are asynchronous.
409		873
410	=item relay => $port, @msg	874	Sending messages in Erlang is synchronous and blocks the process (and
		875	so does not need a queue that can overflow). AEMP sends are immediate,
		876	connection establishment is handled in the background.
411		877
412	Simply forwards the message to the given port.	878	=item * Erlang suffers from silent message loss, AEMP does not.
413		879
414	=cut	880	Erlang implements few guarantees on messages delivery - messages can get
		881	lost without any of the processes realising it (i.e. you send messages a,
		882	b, and c, and the other side only receives messages a and c).
415		883
416	rcv "", relay => sub { \&snd; () };	884	AEMP guarantees correct ordering, and the guarantee that after one message
		885	is lost, all following ones sent to the same port are lost as well, until
		886	monitoring raises an error, so there are no silent "holes" in the message
		887	sequence.
417		888
418	=item eval => $string[ @reply]	889	=item * Erlang can send messages to the wrong port, AEMP does not.
419		890
420	Evaluates the given string. If C<@reply> is given, then a message of the	891	In Erlang it is quite likely that a node that restarts reuses a process ID
421	form C<@reply, $@, @evalres> is sent.	892	known to other nodes for a completely different process, causing messages
		893	destined for that process to end up in an unrelated process.
422		894
423	Example: crash another node.	895	AEMP never reuses port IDs, so old messages or old port IDs floating
		896	around in the network will not be sent to an unrelated port.
424		897
425	snd $othernode, eval => "exit";	898	=item * Erlang uses unprotected connections, AEMP uses secure
		899	authentication and can use TLS.
426		900
427	=cut	901	AEMP can use a proven protocol - TLS - to protect connections and
		902	securely authenticate nodes.
428		903
429	rcv "", eval => sub {	904	=item * The AEMP protocol is optimised for both text-based and binary
430	my (undef, $string, @reply) = @_;	905	communications.
431	my @res = eval $string;
432	snd @reply, "$@", @res if @reply;
433	()
434	};
435		906
436	=item time => @reply	907	The AEMP protocol, unlike the Erlang protocol, supports both programming
		908	language independent text-only protocols (good for debugging) and binary,
		909	language-specific serialisers (e.g. Storable). By default, unless TLS is
		910	used, the protocol is actually completely text-based.
437		911
438	Replies the the current node time to C<@reply>.	912	It has also been carefully designed to be implementable in other languages
		913	with a minimum of work while gracefully degrading functionality to make the
		914	protocol simple.
439		915
440	Example: tell the current node to send the current time to C<$myport> in a	916	=item * AEMP has more flexible monitoring options than Erlang.
441	C<timereply> message.
442		917
443	snd $NODE, time => $myport, timereply => 1, 2;	918	In Erlang, you can chose to receive I<all> exit signals as messages
444	# => snd $myport, timereply => 1, 2, <time>	919	or I<none>, there is no in-between, so monitoring single processes is
		920	difficult to implement. Monitoring in AEMP is more flexible than in
		921	Erlang, as one can choose between automatic kill, exit message or callback
		922	on a per-process basis.
445		923
446	=cut	924	=item * Erlang tries to hide remote/local connections, AEMP does not.
447		925
448	rcv "", time => sub { shift; snd @_, AE::time; () };	926	Monitoring in Erlang is not an indicator of process death/crashes, in the
		927	same way as linking is (except linking is unreliable in Erlang).
		928
		929	In AEMP, you don't "look up" registered port names or send to named ports
		930	that might or might not be persistent. Instead, you normally spawn a port
		931	on the remote node. The init function monitors you, and you monitor the
		932	remote port. Since both monitors are local to the node, they are much more
		933	reliable (no need for C<spawn_link>).
		934
		935	This also saves round-trips and avoids sending messages to the wrong port
		936	(hard to do in Erlang).
449		937
450	=back	938	=back
451		939
		940	=head1 RATIONALE
		941
		942	=over 4
		943
		944	=item Why strings for port and node IDs, why not objects?
		945
		946	We considered "objects", but found that the actual number of methods
		947	that can be called are quite low. Since port and node IDs travel over
		948	the network frequently, the serialising/deserialising would add lots of
		949	overhead, as well as having to keep a proxy object everywhere.
		950
		951	Strings can easily be printed, easily serialised etc. and need no special
		952	procedures to be "valid".
		953
		954	And as a result, a miniport consists of a single closure stored in a
		955	global hash - it can't become much cheaper.
		956
		957	=item Why favour JSON, why not a real serialising format such as Storable?
		958
		959	In fact, any AnyEvent::MP node will happily accept Storable as framing
		960	format, but currently there is no way to make a node use Storable by
		961	default (although all nodes will accept it).
		962
		963	The default framing protocol is JSON because a) JSON::XS is many times
		964	faster for small messages and b) most importantly, after years of
		965	experience we found that object serialisation is causing more problems
		966	than it solves: Just like function calls, objects simply do not travel
		967	easily over the network, mostly because they will always be a copy, so you
		968	always have to re-think your design.
		969
		970	Keeping your messages simple, concentrating on data structures rather than
		971	objects, will keep your messages clean, tidy and efficient.
		972
		973	=back
		974
452	=head1 SEE ALSO	975	=head1 SEE ALSO
		976
		977	L<AnyEvent::MP::Intro> - a gentle introduction.
		978
		979	L<AnyEvent::MP::Kernel> - more, lower-level, stuff.
		980
		981	L<AnyEvent::MP::Global> - network maintainance and port groups, to find
		982	your applications.
		983
		984	L<AnyEvent::MP::LogCatcher> - simple service to display log messages from
		985	all nodes.
453		986
454	L<AnyEvent>.	987	L<AnyEvent>.
455		988
456	=head1 AUTHOR	989	=head1 AUTHOR
457		990

Diff Legend

-–
+Removed lines
-+
+Added lines
-<
+Changed lines
->
+Changed lines

Comparing AnyEvent-MP/MP.pm (file contents): Revision 1.7 by root, Sat Aug 1 15:04:30 2009 UTC vs. Revision 1.104 by root, Fri Nov 6 17:47:20 2009 UTC

Diff Legend

Comparing AnyEvent-MP/MP.pm (file contents):
Revision 1.7 by root, Sat Aug 1 15:04:30 2009 UTC vs.
Revision 1.104 by root, Fri Nov 6 17:47:20 2009 UTC