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

Comparing AnyEvent-MP/MP.pm (file contents):
Revision 1.31 by root, Wed Aug 5 19:55:58 2009 UTC vs.
Revision 1.73 by root, Mon Aug 31 11:08:25 2009 UTC

…		…
8		8
9	$NODE # contains this node's noderef	9	$NODE # contains this node's noderef
10	NODE # returns this node's noderef	10	NODE # returns this node's noderef
11	NODE $port # returns the noderef of the port	11	NODE $port # returns the noderef of the port
12		12
		13	$SELF # receiving/own port id in rcv callbacks
		14
		15	# initialise the node so it can send/receive messages
		16	configure;
		17
		18	# ports are message endpoints
		19
		20	# sending messages
13	snd $port, type => data...;	21	snd $port, type => data...;
		22	snd $port, @msg;
		23	snd @msg_with_first_element_being_a_port;
14		24
15	$SELF # receiving/own port id in rcv callbacks	25	# creating/using ports, the simple way
		26	my $simple_port = port { my @msg = @_ };
16		27
17	rcv $port, smartmatch => $cb->($port, @msg);	28	# creating/using ports, tagged message matching
18		29	my $port = port;
19	# examples:
20	rcv $port2, ping => sub { snd $_[0], "pong"; 0 };	30	rcv $port, ping => sub { snd $_[0], "pong" };
21	rcv $port1, pong => sub { warn "pong received\n" };	31	rcv $port, pong => sub { warn "pong received\n" };
22	snd $port2, ping => $port1;
23		32
24	# more, smarter, matches (_any_ is exported by this module)	33	# create a port on another node
25	rcv $port, [child_died => $pid] => sub { ...	34	my $port = spawn $node, $initfunc, @initdata;
26	rcv $port, [_any_, _any_, 3] => sub { .. $_[2] is 3	35
		36	# monitoring
		37	mon $port, $cb->(@msg) # callback is invoked on death
		38	mon $port, $otherport # kill otherport on abnormal death
		39	mon $port, $otherport, @msg # send message on death
		40
		41	=head1 CURRENT STATUS
		42
		43	bin/aemp - stable.
		44	AnyEvent::MP - stable API, should work.
		45	AnyEvent::MP::Intro - uptodate, but incomplete.
		46	AnyEvent::MP::Kernel - mostly stable.
		47	AnyEvent::MP::Global - stable API, protocol not yet final.
		48
		49	stay tuned.
27		50
28	=head1 DESCRIPTION	51	=head1 DESCRIPTION
29		52
30	This module (-family) implements a simple message passing framework.	53	This module (-family) implements a simple message passing framework.
31		54
32	Despite its simplicity, you can securely message other processes running	55	Despite its simplicity, you can securely message other processes running
33	on the same or other hosts.	56	on the same or other hosts, and you can supervise entities remotely.
34		57
35	For an introduction to this module family, see the L<AnyEvent::MP::Intro>	58	For an introduction to this module family, see the L<AnyEvent::MP::Intro>
36	manual page.	59	manual page and the examples under F<eg/>.
37		60
38	At the moment, this module family is severly broken and underdocumented,	61	At the moment, this module family is a bit underdocumented.
39	so do not use. This was uploaded mainly to reserve the CPAN namespace -
40	stay tuned! The basic API should be finished, however.
41		62
42	=head1 CONCEPTS	63	=head1 CONCEPTS
43		64
44	=over 4	65	=over 4
45		66
46	=item port	67	=item port
47		68
48	A port is something you can send messages to (with the C<snd> function).	69	A port is something you can send messages to (with the C<snd> function).
49		70
50	Some ports allow you to register C<rcv> handlers that can match specific	71	Ports allow you to register C<rcv> handlers that can match all or just
51	messages. All C<rcv> handlers will receive messages they match, messages	72	some messages. Messages send to ports will not be queued, regardless of
52	will not be queued.	73	anything was listening for them or not.
53		74
54	=item port id - C<noderef#portname>	75	=item port ID - C<nodeid#portname>
55		76
56	A port id is normaly the concatenation of a noderef, a hash-mark (C<#>) as	77	A port ID is the concatenation of a node ID, a hash-mark (C<#>) as
57	separator, and a port name (a printable string of unspecified format). An	78	separator, and a port name (a printable string of unspecified format).
58	exception is the the node port, whose ID is identical to its node
59	reference.
60		79
61	=item node	80	=item node
62		81
63	A node is a single process containing at least one port - the node	82	A node is a single process containing at least one port - the node port,
64	port. You can send messages to node ports to find existing ports or to	83	which enables nodes to manage each other remotely, and to create new
65	create new ports, among other things.	84	ports.
66		85
67	Nodes are either private (single-process only), slaves (connected to a	86	Nodes are either public (have one or more listening ports) or private
68	master node only) or public nodes (connectable from unrelated nodes).	87	(no listening ports). Private nodes cannot talk to other private nodes
		88	currently.
69		89
70	=item noderef - C<host:port,host:port...>, C<id@noderef>, C<id>	90	=item node ID - C<[a-za-Z0-9_\-.:]+>
71		91
72	A node reference is a string that either simply identifies the node (for	92	A node ID is a string that uniquely identifies the node within a
73	private and slave nodes), or contains a recipe on how to reach a given	93	network. Depending on the configuration used, node IDs can look like a
74	node (for public nodes).	94	hostname, a hostname and a port, or a random string. AnyEvent::MP itself
		95	doesn't interpret node IDs in any way.
75		96
76	This recipe is simply a comma-separated list of C<address:port> pairs (for	97	=item binds - C<ip:port>
77	TCP/IP, other protocols might look different).
78		98
79	Node references come in two flavours: resolved (containing only numerical	99	Nodes can only talk to each other by creating some kind of connection to
80	addresses) or unresolved (where hostnames are used instead of addresses).	100	each other. To do this, nodes should listen on one or more local transport
		101	endpoints - binds. Currently, only standard C<ip:port> specifications can
		102	be used, which specify TCP ports to listen on.
81		103
82	Before using an unresolved node reference in a message you first have to	104	=item seeds - C<host:port>
83	resolve it.	105
		106	When a node starts, it knows nothing about the network. To teach the node
		107	about the network it first has to contact some other node within the
		108	network. This node is called a seed.
		109
		110	Seeds are transport endpoint(s) of as many nodes as one wants. Those nodes
		111	are expected to be long-running, and at least one of those should always
		112	be available. When nodes run out of connections (e.g. due to a network
		113	error), they try to re-establish connections to some seednodes again to
		114	join the network.
		115
		116	Apart from being sued for seeding, seednodes are not special in any way -
		117	every public node can be a seednode.
84		118
85	=back	119	=back
86		120
87	=head1 VARIABLES/FUNCTIONS	121	=head1 VARIABLES/FUNCTIONS
88		122
…		…
90		124
91	=cut	125	=cut
92		126
93	package AnyEvent::MP;	127	package AnyEvent::MP;
94		128
95	use AnyEvent::MP::Base;	129	use AnyEvent::MP::Kernel;
96		130
97	use common::sense;	131	use common::sense;
98		132
99	use Carp ();	133	use Carp ();
100		134
101	use AE ();	135	use AE ();
102		136
103	use base "Exporter";	137	use base "Exporter";
104		138
105	our $VERSION = '0.1';	139	our $VERSION = $AnyEvent::MP::Kernel::VERSION;
		140
106	our @EXPORT = qw(	141	our @EXPORT = qw(
107	NODE $NODE *SELF node_of _any_	142	NODE $NODE *SELF node_of after
108	resolve_node initialise_node	143	configure
109	snd rcv mon kil reg psub	144	snd rcv mon mon_guard kil reg psub spawn
110	port	145	port
111	);	146	);
112		147
113	our $SELF;	148	our $SELF;
114		149
…		…
118	kil $SELF, die => $msg;	153	kil $SELF, die => $msg;
119	}	154	}
120		155
121	=item $thisnode = NODE / $NODE	156	=item $thisnode = NODE / $NODE
122		157
123	The C<NODE> function returns, and the C<$NODE> variable contains	158	The C<NODE> function returns, and the C<$NODE> variable contains, the node
124	the noderef of the local node. The value is initialised by a call	159	ID of the node running in the current process. This value is initialised by
125	to C<become_public> or C<become_slave>, after which all local port	160	a call to C<configure>.
126	identifiers become invalid.
127		161
128	=item $noderef = node_of $portid	162	=item $nodeid = node_of $port
129		163
130	Extracts and returns the noderef from a portid or a noderef.	164	Extracts and returns the node ID from a port ID or a node ID.
131		165
132	=item $cv = resolve_node $noderef	166	=item configure key => value...
133		167
134	Takes an unresolved node reference that may contain hostnames and	168	Before a node can talk to other nodes on the network (i.e. enter
135	abbreviated IDs, resolves all of them and returns a resolved node	169	"distributed mode") it has to configure itself - the minimum a node needs
136	reference.	170	to know is its own name, and optionally it should know the addresses of
		171	some other nodes in the network to discover other nodes.
137		172
138	In addition to C<address:port> pairs allowed in resolved noderefs, the	173	This function configures a node - it must be called exactly once (or
139	following forms are supported:	174	never) before calling other AnyEvent::MP functions.
140		175
141	=over 4	176	=over 4
142		177
143	=item the empty string	178	=item step 1, gathering configuration from profiles
144		179
145	An empty-string component gets resolved as if the default port (4040) was	180	The function first looks up a profile in the aemp configuration (see the
146	specified.	181	L<aemp> commandline utility). The profile name can be specified via the
		182	named C<profile> parameter. If it is missing, then the nodename (F<uname
		183	-n>) will be used as profile name.
147		184
148	=item naked port numbers (e.g. C<1234>)	185	The profile data is then gathered as follows:
149		186
150	These are resolved by prepending the local nodename and a colon, to be	187	First, all remaining key => value pairs (all of which are conviniently
151	further resolved.	188	undocumented at the moment) will be interpreted as configuration
		189	data. Then they will be overwritten by any values specified in the global
		190	default configuration (see the F<aemp> utility), then the chain of
		191	profiles chosen by the profile name (and any C<parent> attributes).
152		192
153	=item hostnames (e.g. C<localhost:1234>, C<localhost>)	193	That means that the values specified in the profile have highest priority
		194	and the values specified directly via C<configure> have lowest priority,
		195	and can only be used to specify defaults.
154		196
155	These are resolved by using AnyEvent::DNS to resolve them, optionally	197	If the profile specifies a node ID, then this will become the node ID of
156	looking up SRV records for the C<aemp=4040> port, if no port was	198	this process. If not, then the profile name will be used as node ID. The
157	specified.	199	special node ID of C<anon/> will be replaced by a random node ID.
		200
		201	=item step 2, bind listener sockets
		202
		203	The next step is to look up the binds in the profile, followed by binding
		204	aemp protocol listeners on all binds specified (it is possible and valid
		205	to have no binds, meaning that the node cannot be contacted form the
		206	outside. This means the node cannot talk to other nodes that also have no
		207	binds, but it can still talk to all "normal" nodes).
		208
		209	If the profile does not specify a binds list, then a default of C<*> is
		210	used, meaning the node will bind on a dynamically-assigned port on every
		211	local IP address it finds.
		212
		213	=item step 3, connect to seed nodes
		214
		215	As the last step, the seeds list from the profile is passed to the
		216	L<AnyEvent::MP::Global> module, which will then use it to keep
		217	connectivity with at least one node at any point in time.
158		218
159	=back	219	=back
		220
		221	Example: become a distributed node using the locla node name as profile.
		222	This should be the most common form of invocation for "daemon"-type nodes.
		223
		224	configure
		225
		226	Example: become an anonymous node. This form is often used for commandline
		227	clients.
		228
		229	configure nodeid => "anon/";
		230
		231	Example: configure a node using a profile called seed, which si suitable
		232	for a seed node as it binds on all local addresses on a fixed port (4040,
		233	customary for aemp).
		234
		235	# use the aemp commandline utility
		236	# aemp profile seed setnodeid anon/ setbinds '*:4040'
		237
		238	# then use it
		239	configure profile => "seed";
		240
		241	# or simply use aemp from the shell again:
		242	# aemp run profile seed
		243
		244	# or provide a nicer-to-remember nodeid
		245	# aemp run profile seed nodeid "$(hostname)"
160		246
161	=item $SELF	247	=item $SELF
162		248
163	Contains the current port id while executing C<rcv> callbacks or C<psub>	249	Contains the current port id while executing C<rcv> callbacks or C<psub>
164	blocks.	250	blocks.
165		251
166	=item SELF, %SELF, @SELF...	252	=item *SELF, SELF, %SELF, @SELF...
167		253
168	Due to some quirks in how perl exports variables, it is impossible to	254	Due to some quirks in how perl exports variables, it is impossible to
169	just export C<$SELF>, all the symbols called C<SELF> are exported by this	255	just export C<$SELF>, all the symbols named C<SELF> are exported by this
170	module, but only C<$SELF> is currently used.	256	module, but only C<$SELF> is currently used.
171		257
172	=item snd $portid, type => @data	258	=item snd $port, type => @data
173		259
174	=item snd $portid, @msg	260	=item snd $port, @msg
175		261
176	Send the given message to the given port ID, which can identify either	262	Send the given message to the given port, which can identify either a
177	a local or a remote port, and can be either a string or soemthignt hat	263	local or a remote port, and must be a port ID.
178	stringifies a sa port ID (such as a port object :).
179		264
180	While the message can be about anything, it is highly recommended to use a	265	While the message can be almost anything, it is highly recommended to
181	string as first element (a portid, or some word that indicates a request	266	use a string as first element (a port ID, or some word that indicates a
182	type etc.).	267	request type etc.) and to consist if only simple perl values (scalars,
		268	arrays, hashes) - if you think you need to pass an object, think again.
183		269
184	The message data effectively becomes read-only after a call to this	270	The message data logically becomes read-only after a call to this
185	function: modifying any argument is not allowed and can cause many	271	function: modifying any argument (or values referenced by them) is
186	problems.	272	forbidden, as there can be considerable time between the call to C<snd>
		273	and the time the message is actually being serialised - in fact, it might
		274	never be copied as within the same process it is simply handed to the
		275	receiving port.
187		276
188	The type of data you can transfer depends on the transport protocol: when	277	The type of data you can transfer depends on the transport protocol: when
189	JSON is used, then only strings, numbers and arrays and hashes consisting	278	JSON is used, then only strings, numbers and arrays and hashes consisting
190	of those are allowed (no objects). When Storable is used, then anything	279	of those are allowed (no objects). When Storable is used, then anything
191	that Storable can serialise and deserialise is allowed, and for the local	280	that Storable can serialise and deserialise is allowed, and for the local
192	node, anything can be passed.	281	node, anything can be passed. Best rely only on the common denominator of
		282	these.
193		283
194	=item kil $portid[, @reason]	284	=item $local_port = port
195		285
196	Kill the specified port with the given C<@reason>.	286	Create a new local port object and returns its port ID. Initially it has
		287	no callbacks set and will throw an error when it receives messages.
197		288
198	If no C<@reason> is specified, then the port is killed "normally" (linked	289	=item $local_port = port { my @msg = @_ }
199	ports will not be kileld, or even notified).
200		290
201	Otherwise, linked ports get killed with the same reason (second form of	291	Creates a new local port, and returns its ID. Semantically the same as
202	C<mon>, see below).	292	creating a port and calling C<rcv $port, $callback> on it.
203		293
204	Runtime errors while evaluating C<rcv> callbacks or inside C<psub> blocks	294	The block will be called for every message received on the port, with the
205	will be reported as reason C<< die => $@ >>.	295	global variable C<$SELF> set to the port ID. Runtime errors will cause the
		296	port to be C<kil>ed. The message will be passed as-is, no extra argument
		297	(i.e. no port ID) will be passed to the callback.
206		298
207	Transport/communication errors are reported as C<< transport_error =>	299	If you want to stop/destroy the port, simply C<kil> it:
208	$message >>.
209		300
210	=item $guard = mon $portid, $cb->(@reason)	301	my $port = port {
		302	my @msg = @_;
		303	...
		304	kil $SELF;
		305	};
211		306
212	=item $guard = mon $portid, $otherport	307	=cut
213		308
214	=item $guard = mon $portid, $otherport, @msg	309	sub rcv($@);
215		310
		311	sub _kilme {
		312	die "received message on port without callback";
		313	}
		314
		315	sub port(;&) {
		316	my $id = "$UNIQ." . $ID++;
		317	my $port = "$NODE#$id";
		318
		319	rcv $port, shift \|\| \&_kilme;
		320
		321	$port
		322	}
		323
		324	=item rcv $local_port, $callback->(@msg)
		325
		326	Replaces the default callback on the specified port. There is no way to
		327	remove the default callback: use C<sub { }> to disable it, or better
		328	C<kil> the port when it is no longer needed.
		329
		330	The global C<$SELF> (exported by this module) contains C<$port> while
		331	executing the callback. Runtime errors during callback execution will
		332	result in the port being C<kil>ed.
		333
		334	The default callback received all messages not matched by a more specific
		335	C<tag> match.
		336
		337	=item rcv $local_port, tag => $callback->(@msg_without_tag), ...
		338
		339	Register (or replace) callbacks to be called on messages starting with the
		340	given tag on the given port (and return the port), or unregister it (when
		341	C<$callback> is C<$undef> or missing). There can only be one callback
		342	registered for each tag.
		343
		344	The original message will be passed to the callback, after the first
		345	element (the tag) has been removed. The callback will use the same
		346	environment as the default callback (see above).
		347
		348	Example: create a port and bind receivers on it in one go.
		349
		350	my $port = rcv port,
		351	msg1 => sub { ... },
		352	msg2 => sub { ... },
		353	;
		354
		355	Example: create a port, bind receivers and send it in a message elsewhere
		356	in one go:
		357
		358	snd $otherport, reply =>
		359	rcv port,
		360	msg1 => sub { ... },
		361	...
		362	;
		363
		364	Example: temporarily register a rcv callback for a tag matching some port
		365	(e.g. for a rpc reply) and unregister it after a message was received.
		366
		367	rcv $port, $otherport => sub {
		368	my @reply = @_;
		369
		370	rcv $SELF, $otherport;
		371	};
		372
		373	=cut
		374
		375	sub rcv($@) {
		376	my $port = shift;
		377	my ($noderef, $portid) = split /#/, $port, 2;
		378
		379	$NODE{$noderef} == $NODE{""}
		380	or Carp::croak "$port: rcv can only be called on local ports, caught";
		381
		382	while (@_) {
		383	if (ref $_[0]) {
		384	if (my $self = $PORT_DATA{$portid}) {
		385	"AnyEvent::MP::Port" eq ref $self
		386	or Carp::croak "$port: rcv can only be called on message matching ports, caught";
		387
		388	$self->[2] = shift;
		389	} else {
		390	my $cb = shift;
		391	$PORT{$portid} = sub {
		392	local $SELF = $port;
		393	eval { &$cb }; _self_die if $@;
		394	};
		395	}
		396	} elsif (defined $_[0]) {
		397	my $self = $PORT_DATA{$portid} \|\|= do {
		398	my $self = bless [$PORT{$port} \|\| sub { }, { }, $port], "AnyEvent::MP::Port";
		399
		400	$PORT{$portid} = sub {
		401	local $SELF = $port;
		402
		403	if (my $cb = $self->[1]{$_[0]}) {
		404	shift;
		405	eval { &$cb }; _self_die if $@;
		406	} else {
		407	&{ $self->[0] };
		408	}
		409	};
		410
		411	$self
		412	};
		413
		414	"AnyEvent::MP::Port" eq ref $self
		415	or Carp::croak "$port: rcv can only be called on message matching ports, caught";
		416
		417	my ($tag, $cb) = splice @_, 0, 2;
		418
		419	if (defined $cb) {
		420	$self->[1]{$tag} = $cb;
		421	} else {
		422	delete $self->[1]{$tag};
		423	}
		424	}
		425	}
		426
		427	$port
		428	}
		429
		430	=item $closure = psub { BLOCK }
		431
		432	Remembers C<$SELF> and creates a closure out of the BLOCK. When the
		433	closure is executed, sets up the environment in the same way as in C<rcv>
		434	callbacks, i.e. runtime errors will cause the port to get C<kil>ed.
		435
		436	This is useful when you register callbacks from C<rcv> callbacks:
		437
		438	rcv delayed_reply => sub {
		439	my ($delay, @reply) = @_;
		440	my $timer = AE::timer $delay, 0, psub {
		441	snd @reply, $SELF;
		442	};
		443	};
		444
		445	=cut
		446
		447	sub psub(&) {
		448	my $cb = shift;
		449
		450	my $port = $SELF
		451	or Carp::croak "psub can only be called from within rcv or psub callbacks, not";
		452
		453	sub {
		454	local $SELF = $port;
		455
		456	if (wantarray) {
		457	my @res = eval { &$cb };
		458	_self_die if $@;
		459	@res
		460	} else {
		461	my $res = eval { &$cb };
		462	_self_die if $@;
		463	$res
		464	}
		465	}
		466	}
		467
		468	=item $guard = mon $port, $cb->(@reason) # call $cb when $port dies
		469
		470	=item $guard = mon $port, $rcvport # kill $rcvport when $port dies
		471
		472	=item $guard = mon $port # kill $SELF when $port dies
		473
		474	=item $guard = mon $port, $rcvport, @msg # send a message when $port dies
		475
216	Monitor the given port and do something when the port is killed.	476	Monitor the given port and do something when the port is killed or
		477	messages to it were lost, and optionally return a guard that can be used
		478	to stop monitoring again.
217		479
		480	C<mon> effectively guarantees that, in the absence of hardware failures,
		481	after starting the monitor, either all messages sent to the port will
		482	arrive, or the monitoring action will be invoked after possible message
		483	loss has been detected. No messages will be lost "in between" (after
		484	the first lost message no further messages will be received by the
		485	port). After the monitoring action was invoked, further messages might get
		486	delivered again.
		487
		488	Note that monitoring-actions are one-shot: once messages are lost (and a
		489	monitoring alert was raised), they are removed and will not trigger again.
		490
218	In the first form, the callback is simply called with any number	491	In the first form (callback), the callback is simply called with any
219	of C<@reason> elements (no @reason means that the port was deleted	492	number of C<@reason> elements (no @reason means that the port was deleted
220	"normally"). Note also that I<< the callback B<must> never die >>, so use	493	"normally"). Note also that I<< the callback B<must> never die >>, so use
221	C<eval> if unsure.	494	C<eval> if unsure.
222		495
223	In the second form, the other port will be C<kil>'ed with C<@reason>, iff	496	In the second form (another port given), the other port (C<$rcvport>)
224	a @reason was specified, i.e. on "normal" kils nothing happens, while	497	will be C<kil>'ed with C<@reason>, iff a @reason was specified, i.e. on
225	under all other conditions, the other port is killed with the same reason.	498	"normal" kils nothing happens, while under all other conditions, the other
		499	port is killed with the same reason.
226		500
		501	The third form (kill self) is the same as the second form, except that
		502	C<$rvport> defaults to C<$SELF>.
		503
227	In the last form, a message of the form C<@msg, @reason> will be C<snd>.	504	In the last form (message), a message of the form C<@msg, @reason> will be
		505	C<snd>.
		506
		507	As a rule of thumb, monitoring requests should always monitor a port from
		508	a local port (or callback). The reason is that kill messages might get
		509	lost, just like any other message. Another less obvious reason is that
		510	even monitoring requests can get lost (for exmaple, when the connection
		511	to the other node goes down permanently). When monitoring a port locally
		512	these problems do not exist.
228		513
229	Example: call a given callback when C<$port> is killed.	514	Example: call a given callback when C<$port> is killed.
230		515
231	mon $port, sub { warn "port died because of <@_>\n" };	516	mon $port, sub { warn "port died because of <@_>\n" };
232		517
233	Example: kill ourselves when C<$port> is killed abnormally.	518	Example: kill ourselves when C<$port> is killed abnormally.
234		519
235	mon $port, $self;	520	mon $port;
236		521
237	Example: send us a restart message another C<$port> is killed.	522	Example: send us a restart message when another C<$port> is killed.
238		523
239	mon $port, $self => "restart";	524	mon $port, $self => "restart";
240		525
241	=cut	526	=cut
242		527
243	sub mon {	528	sub mon {
244	my ($noderef, $port) = split /#/, shift, 2;	529	my ($noderef, $port) = split /#/, shift, 2;
245		530
246	my $node = $NODE{$noderef} \|\| add_node $noderef;	531	my $node = $NODE{$noderef} \|\| add_node $noderef;
247		532
248	my $cb = shift;	533	my $cb = @_ ? shift : $SELF \|\| Carp::croak 'mon: called with one argument only, but $SELF not set,';
249		534
250	unless (ref $cb) {	535	unless (ref $cb) {
251	if (@_) {	536	if (@_) {
252	# send a kill info message	537	# send a kill info message
253	my (@msg) = ($cb, @_);	538	my (@msg) = ($cb, @_);
…		…
271	is killed, the references will be freed.	556	is killed, the references will be freed.
272		557
273	Optionally returns a guard that will stop the monitoring.	558	Optionally returns a guard that will stop the monitoring.
274		559
275	This function is useful when you create e.g. timers or other watchers and	560	This function is useful when you create e.g. timers or other watchers and
276	want to free them when the port gets killed:	561	want to free them when the port gets killed (note the use of C<psub>):
277		562
278	$port->rcv (start => sub {	563	$port->rcv (start => sub {
279	my $timer; $timer = mon_guard $port, AE::timer 1, 1, sub {	564	my $timer; $timer = mon_guard $port, AE::timer 1, 1, psub {
280	undef $timer if 0.9 < rand;	565	undef $timer if 0.9 < rand;
281	});	566	});
282	});	567	});
283		568
284	=cut	569	=cut
285		570
286	sub mon_guard {	571	sub mon_guard {
287	my ($port, @refs) = @_;	572	my ($port, @refs) = @_;
288		573
		574	#TODO: mon-less form?
		575
289	mon $port, sub { 0 && @refs }	576	mon $port, sub { 0 && @refs }
290	}	577	}
291		578
292	=item lnk $port1, $port2	579	=item kil $port[, @reason]
293		580
294	Link two ports. This is simply a shorthand for:	581	Kill the specified port with the given C<@reason>.
295		582
296	mon $port1, $port2;	583	If no C<@reason> is specified, then the port is killed "normally" (ports
297	mon $port2, $port1;	584	monitoring other ports will not necessarily die because a port dies
		585	"normally").
298		586
299	It means that if either one is killed abnormally, the other one gets	587	Otherwise, linked ports get killed with the same reason (second form of
300	killed as well.	588	C<mon>, see above).
301		589
302	=item $local_port = port	590	Runtime errors while evaluating C<rcv> callbacks or inside C<psub> blocks
		591	will be reported as reason C<< die => $@ >>.
303		592
304	Create a new local port object that can be used either as a pattern	593	Transport/communication errors are reported as C<< transport_error =>
305	matching port ("full port") or a single-callback port ("miniport"),	594	$message >>.
306	depending on how C<rcv> callbacks are bound to the object.
307		595
308	=item $portid = port { my @msg = @_; $finished }
309
310	Creates a "mini port", that is, a very lightweight port without any
311	pattern matching behind it, and returns its ID.
312
313	The block will be called for every message received on the port. When the
314	callback returns a true value its job is considered "done" and the port
315	will be destroyed. Otherwise it will stay alive.
316
317	The message will be passed as-is, no extra argument (i.e. no port id) will
318	be passed to the callback.
319
320	If you need the local port id in the callback, this works nicely:
321
322	my $port; $port = port {
323	snd $otherport, reply => $port;
324	};
325
326	=cut	596	=cut
327		597
328	sub port(;&) {	598	=item $port = spawn $node, $initfunc[, @initdata]
329	my $id = "$UNIQ." . $ID++;
330	my $port = "$NODE#$id";
331		599
332	if (@_) {	600	Creates a port on the node C<$node> (which can also be a port ID, in which
333	my $cb = shift;	601	case it's the node where that port resides).
334	$PORT{$id} = sub {	602
335	local $SELF = $port;	603	The port ID of the newly created port is returned immediately, and it is
336	eval {	604	possible to immediately start sending messages or to monitor the port.
337	&$cb	605
338	and kil $id;	606	After the port has been created, the init function is called on the remote
		607	node, in the same context as a C<rcv> callback. This function must be a
		608	fully-qualified function name (e.g. C<MyApp::Chat::Server::init>). To
		609	specify a function in the main program, use C<::name>.
		610
		611	If the function doesn't exist, then the node tries to C<require>
		612	the package, then the package above the package and so on (e.g.
		613	C<MyApp::Chat::Server>, C<MyApp::Chat>, C<MyApp>) until the function
		614	exists or it runs out of package names.
		615
		616	The init function is then called with the newly-created port as context
		617	object (C<$SELF>) and the C<@initdata> values as arguments.
		618
		619	A common idiom is to pass a local port, immediately monitor the spawned
		620	port, and in the remote init function, immediately monitor the passed
		621	local port. This two-way monitoring ensures that both ports get cleaned up
		622	when there is a problem.
		623
		624	Example: spawn a chat server port on C<$othernode>.
		625
		626	# this node, executed from within a port context:
		627	my $server = spawn $othernode, "MyApp::Chat::Server::connect", $SELF;
		628	mon $server;
		629
		630	# init function on C<$othernode>
		631	sub connect {
		632	my ($srcport) = @_;
		633
		634	mon $srcport;
		635
		636	rcv $SELF, sub {
339	};	637	...
340	_self_die if $@;
341	};
342	} else {
343	my $self = bless {
344	id => "$NODE#$id",
345	}, "AnyEvent::MP::Port";
346
347	$PORT_DATA{$id} = $self;
348	$PORT{$id} = sub {
349	local $SELF = $port;
350
351	eval {
352	for (@{ $self->{rc0}{$_[0]} }) {
353	$_ && &{$_->[0]}
354	&& undef $_;
355	}
356
357	for (@{ $self->{rcv}{$_[0]} }) {
358	$_ && [@_[1 .. @{$_->[1]}]] ~~ $_->[1]
359	&& &{$_->[0]}
360	&& undef $_;
361	}
362
363	for (@{ $self->{any} }) {
364	$_ && [@_[0 .. $#{$_->[1]}]] ~~ $_->[1]
365	&& &{$_->[0]}
366	&& undef $_;
367	}
368	};
369	_self_die if $@;
370	};	638	};
371	}	639	}
372		640
373	$port
374	}
375
376	=item reg $portid, $name
377
378	Registers the given port under the name C<$name>. If the name already
379	exists it is replaced.
380
381	A port can only be registered under one well known name.
382
383	A port automatically becomes unregistered when it is killed.
384
385	=cut	641	=cut
386		642
387	sub reg(@) {	643	sub _spawn {
388	my ($portid, $name) = @_;
389
390	$REG{$name} = $portid;
391	}
392
393	=item rcv $portid, $callback->(@msg)
394
395	Replaces the callback on the specified miniport (or newly created port
396	object, see C<port>). Full ports are configured with the following calls:
397
398	=item rcv $portid, tagstring => $callback->(@msg), ...
399
400	=item rcv $portid, $smartmatch => $callback->(@msg), ...
401
402	=item rcv $portid, [$smartmatch...] => $callback->(@msg), ...
403
404	Register callbacks to be called on matching messages on the given port.
405
406	The callback has to return a true value when its work is done, after
407	which is will be removed, or a false value in which case it will stay
408	registered.
409
410	The global C<$SELF> (exported by this module) contains C<$portid> while
411	executing the callback.
412
413	Runtime errors wdurign callback execution will result in the port being
414	C<kil>ed.
415
416	If the match is an array reference, then it will be matched against the
417	first elements of the message, otherwise only the first element is being
418	matched.
419
420	Any element in the match that is specified as C<_any_> (a function
421	exported by this module) matches any single element of the message.
422
423	While not required, it is highly recommended that the first matching
424	element is a string identifying the message. The one-string-only match is
425	also the most efficient match (by far).
426
427	=cut
428
429	sub rcv($@) {
430	my $portid = shift;	644	my $port = shift;
431	my ($noderef, $port) = split /#/, $port, 2;	645	my $init = shift;
432		646
433	($NODE{$noderef} \|\| add_node $noderef) == $NODE{""}	647	local $SELF = "$NODE#$port";
434	or Carp::croak "$noderef#$port: rcv can only be called on local ports, caught";	648	eval {
435		649	&{ load_func $init }
436	my $self = $PORT_DATA{$port}
437	or Carp::croak "$noderef#$port: rcv can only be called on message matching ports, caught";
438
439	"AnyEvent::MP::Port" eq ref $self
440	or Carp::croak "$noderef#$port: rcv can only be called on message matching ports, caught";
441
442	while (@_) {
443	my ($match, $cb) = splice @_, 0, 2;
444
445	if (!ref $match) {
446	push @{ $self->{rc0}{$match} }, [$cb];
447	} elsif (("ARRAY" eq ref $match && !ref $match->[0])) {
448	my ($type, @match) = @$match;
449	@match
450	? push @{ $self->{rcv}{$match->[0]} }, [$cb, \@match]
451	: push @{ $self->{rc0}{$match->[0]} }, [$cb];
452	} else {
453	push @{ $self->{any} }, [$cb, $match];
454	}
455	}
456
457	$portid
458	}
459
460	=item $closure = psub { BLOCK }
461
462	Remembers C<$SELF> and creates a closure out of the BLOCK. When the
463	closure is executed, sets up the environment in the same way as in C<rcv>
464	callbacks, i.e. runtime errors will cause the port to get C<kil>ed.
465
466	This is useful when you register callbacks from C<rcv> callbacks:
467
468	rcv delayed_reply => sub {
469	my ($delay, @reply) = @_;
470	my $timer = AE::timer $delay, 0, psub {
471	snd @reply, $SELF;
472	};
473	};	650	};
474
475	=cut
476
477	sub psub(&) {
478	my $cb = shift;
479
480	my $port = $SELF
481	or Carp::croak "psub can only be called from within rcv or psub callbacks, not";
482
483	sub {
484	local $SELF = $port;
485
486	if (wantarray) {
487	my @res = eval { &$cb };
488	_self_die if $@;	651	_self_die if $@;
489	@res	652	}
490	} else {	653
491	my $res = eval { &$cb };	654	sub spawn(@) {
492	_self_die if $@;	655	my ($noderef, undef) = split /#/, shift, 2;
493	$res	656
494	}	657	my $id = "$RUNIQ." . $ID++;
		658
		659	$_[0] =~ /::/
		660	or Carp::croak "spawn init function must be a fully-qualified name, caught";
		661
		662	snd_to_func $noderef, "AnyEvent::MP::_spawn" => $id, @_;
		663
		664	"$noderef#$id"
		665	}
		666
		667	=item after $timeout, @msg
		668
		669	=item after $timeout, $callback
		670
		671	Either sends the given message, or call the given callback, after the
		672	specified number of seconds.
		673
		674	This is simply a utility function that comes in handy at times - the
		675	AnyEvent::MP author is not convinced of the wisdom of having it, though,
		676	so it may go away in the future.
		677
		678	=cut
		679
		680	sub after($@) {
		681	my ($timeout, @action) = @_;
		682
		683	my $t; $t = AE::timer $timeout, 0, sub {
		684	undef $t;
		685	ref $action[0]
		686	? $action[0]()
		687	: snd @action;
495	}	688	};
496	}	689	}
497		690
498	=back	691	=back
499		692
500	=head1 FUNCTIONS FOR NODES	693	=head1 AnyEvent::MP vs. Distributed Erlang
		694
		695	AnyEvent::MP got lots of its ideas from distributed Erlang (Erlang node
		696	== aemp node, Erlang process == aemp port), so many of the documents and
		697	programming techniques employed by Erlang apply to AnyEvent::MP. Here is a
		698	sample:
		699
		700	http://www.Erlang.se/doc/programming_rules.shtml
		701	http://Erlang.org/doc/getting_started/part_frame.html # chapters 3 and 4
		702	http://Erlang.org/download/Erlang-book-part1.pdf # chapters 5 and 6
		703	http://Erlang.org/download/armstrong_thesis_2003.pdf # chapters 4 and 5
		704
		705	Despite the similarities, there are also some important differences:
501		706
502	=over 4	707	=over 4
503		708
504	=item become_public $noderef	709	=item * Node IDs are arbitrary strings in AEMP.
505		710
506	Tells the node to become a public node, i.e. reachable from other nodes.
507
508	The first argument is the (unresolved) node reference of the local node
509	(if missing then the empty string is used).
510
511	It is quite common to not specify anything, in which case the local node
512	tries to listen on the default port, or to only specify a port number, in
513	which case AnyEvent::MP tries to guess the local addresses.
514
515	=cut
516
517	=back
518
519	=head1 NODE MESSAGES
520
521	Nodes understand the following messages sent to them. Many of them take
522	arguments called C<@reply>, which will simply be used to compose a reply
523	message - C<$reply[0]> is the port to reply to, C<$reply[1]> the type and
524	the remaining arguments are simply the message data.
525
526	While other messages exist, they are not public and subject to change.
527
528	=over 4
529
530	=cut
531
532	=item lookup => $name, @reply
533
534	Replies with the port ID of the specified well-known port, or C<undef>.
535
536	=item devnull => ...
537
538	Generic data sink/CPU heat conversion.
539
540	=item relay => $port, @msg
541
542	Simply forwards the message to the given port.
543
544	=item eval => $string[ @reply]
545
546	Evaluates the given string. If C<@reply> is given, then a message of the
547	form C<@reply, $@, @evalres> is sent.
548
549	Example: crash another node.
550
551	snd $othernode, eval => "exit";
552
553	=item time => @reply
554
555	Replies the the current node time to C<@reply>.
556
557	Example: tell the current node to send the current time to C<$myport> in a
558	C<timereply> message.
559
560	snd $NODE, time => $myport, timereply => 1, 2;
561	# => snd $myport, timereply => 1, 2, <time>
562
563	=back
564
565	=head1 AnyEvent::MP vs. Distributed Erlang
566
567	AnyEvent::MP got lots of its ideas from distributed erlang (erlang node
568	== aemp node, erlang process == aemp port), so many of the documents and
569	programming techniques employed by erlang apply to AnyEvent::MP. Here is a
570	sample:
571
572	http://www.erlang.se/doc/programming_rules.shtml
573	http://erlang.org/doc/getting_started/part_frame.html # chapters 3 and 4
574	http://erlang.org/download/erlang-book-part1.pdf # chapters 5 and 6
575	http://erlang.org/download/armstrong_thesis_2003.pdf # chapters 4 and 5
576
577	Despite the similarities, there are also some important differences:
578
579	=over 4
580
581	=item * Node references contain the recipe on how to contact them.
582
583	Erlang relies on special naming and DNS to work everywhere in the	711	Erlang relies on special naming and DNS to work everywhere in the same
584	same way. AEMP relies on each node knowing it's own address(es), with	712	way. AEMP relies on each node somehow knowing its own address(es) (e.g. by
585	convenience functionality.	713	configuraiton or DNS), but will otherwise discover other odes itself.
586		714
587	This means that AEMP requires a less tightly controlled environment at the	715	=item * Erlang has a "remote ports are like local ports" philosophy, AEMP
588	cost of longer node references and a slightly higher management overhead.	716	uses "local ports are like remote ports".
		717
		718	The failure modes for local ports are quite different (runtime errors
		719	only) then for remote ports - when a local port dies, you I<know> it dies,
		720	when a connection to another node dies, you know nothing about the other
		721	port.
		722
		723	Erlang pretends remote ports are as reliable as local ports, even when
		724	they are not.
		725
		726	AEMP encourages a "treat remote ports differently" philosophy, with local
		727	ports being the special case/exception, where transport errors cannot
		728	occur.
589		729
590	=item * Erlang uses processes and a mailbox, AEMP does not queue.	730	=item * Erlang uses processes and a mailbox, AEMP does not queue.
591		731
592	Erlang uses processes that selctively receive messages, and therefore	732	Erlang uses processes that selectively receive messages, and therefore
593	needs a queue. AEMP is event based, queuing messages would serve no useful	733	needs a queue. AEMP is event based, queuing messages would serve no
594	purpose.	734	useful purpose. For the same reason the pattern-matching abilities of
		735	AnyEvent::MP are more limited, as there is little need to be able to
		736	filter messages without dequeing them.
595		737
596	(But see L<Coro::MP> for a more erlang-like process model on top of AEMP).	738	(But see L<Coro::MP> for a more Erlang-like process model on top of AEMP).
597		739
598	=item * Erlang sends are synchronous, AEMP sends are asynchronous.	740	=item * Erlang sends are synchronous, AEMP sends are asynchronous.
599		741
600	Sending messages in erlang is synchronous and blocks the process. AEMP	742	Sending messages in Erlang is synchronous and blocks the process (and
601	sends are immediate, connection establishment is handled in the	743	so does not need a queue that can overflow). AEMP sends are immediate,
602	background.	744	connection establishment is handled in the background.
603		745
604	=item * Erlang can silently lose messages, AEMP cannot.	746	=item * Erlang suffers from silent message loss, AEMP does not.
605		747
606	Erlang makes few guarantees on messages delivery - messages can get lost	748	Erlang makes few guarantees on messages delivery - messages can get lost
607	without any of the processes realising it (i.e. you send messages a, b,	749	without any of the processes realising it (i.e. you send messages a, b,
608	and c, and the other side only receives messages a and c).	750	and c, and the other side only receives messages a and c).
609		751
610	AEMP guarantees correct ordering, and the guarantee that there are no	752	AEMP guarantees correct ordering, and the guarantee that after one message
611	holes in the message sequence.	753	is lost, all following ones sent to the same port are lost as well, until
612		754	monitoring raises an error, so there are no silent "holes" in the message
613	=item * In erlang, processes can be declared dead and later be found to be	755	sequence.
614	alive.
615
616	In erlang it can happen that a monitored process is declared dead and
617	linked processes get killed, but later it turns out that the process is
618	still alive - and can receive messages.
619
620	In AEMP, when port monitoring detects a port as dead, then that port will
621	eventually be killed - it cannot happen that a node detects a port as dead
622	and then later sends messages to it, finding it is still alive.
623		756
624	=item * Erlang can send messages to the wrong port, AEMP does not.	757	=item * Erlang can send messages to the wrong port, AEMP does not.
625		758
626	In erlang it is quite possible that a node that restarts reuses a process	759	In Erlang it is quite likely that a node that restarts reuses a process ID
627	ID known to other nodes for a completely different process, causing	760	known to other nodes for a completely different process, causing messages
628	messages destined for that process to end up in an unrelated process.	761	destined for that process to end up in an unrelated process.
629		762
630	AEMP never reuses port IDs, so old messages or old port IDs floating	763	AEMP never reuses port IDs, so old messages or old port IDs floating
631	around in the network will not be sent to an unrelated port.	764	around in the network will not be sent to an unrelated port.
632		765
633	=item * Erlang uses unprotected connections, AEMP uses secure	766	=item * Erlang uses unprotected connections, AEMP uses secure
634	authentication and can use TLS.	767	authentication and can use TLS.
635		768
636	AEMP can use a proven protocol - SSL/TLS - to protect connections and	769	AEMP can use a proven protocol - TLS - to protect connections and
637	securely authenticate nodes.	770	securely authenticate nodes.
638		771
639	=item * The AEMP protocol is optimised for both text-based and binary	772	=item * The AEMP protocol is optimised for both text-based and binary
640	communications.	773	communications.
641		774
642	The AEMP protocol, unlike the erlang protocol, supports both	775	The AEMP protocol, unlike the Erlang protocol, supports both programming
643	language-independent text-only protocols (good for debugging) and binary,	776	language independent text-only protocols (good for debugging) and binary,
644	language-specific serialisers (e.g. Storable).	777	language-specific serialisers (e.g. Storable). By default, unless TLS is
		778	used, the protocol is actually completely text-based.
645		779
646	It has also been carefully designed to be implementable in other languages	780	It has also been carefully designed to be implementable in other languages
647	with a minimum of work while gracefully degrading fucntionality to make the	781	with a minimum of work while gracefully degrading functionality to make the
648	protocol simple.	782	protocol simple.
649		783
		784	=item * AEMP has more flexible monitoring options than Erlang.
		785
		786	In Erlang, you can chose to receive I<all> exit signals as messages
		787	or I<none>, there is no in-between, so monitoring single processes is
		788	difficult to implement. Monitoring in AEMP is more flexible than in
		789	Erlang, as one can choose between automatic kill, exit message or callback
		790	on a per-process basis.
		791
		792	=item * Erlang tries to hide remote/local connections, AEMP does not.
		793
		794	Monitoring in Erlang is not an indicator of process death/crashes, in the
		795	same way as linking is (except linking is unreliable in Erlang).
		796
		797	In AEMP, you don't "look up" registered port names or send to named ports
		798	that might or might not be persistent. Instead, you normally spawn a port
		799	on the remote node. The init function monitors you, and you monitor the
		800	remote port. Since both monitors are local to the node, they are much more
		801	reliable (no need for C<spawn_link>).
		802
		803	This also saves round-trips and avoids sending messages to the wrong port
		804	(hard to do in Erlang).
		805
650	=back	806	=back
651		807
		808	=head1 RATIONALE
		809
		810	=over 4
		811
		812	=item Why strings for port and node IDs, why not objects?
		813
		814	We considered "objects", but found that the actual number of methods
		815	that can be called are quite low. Since port and node IDs travel over
		816	the network frequently, the serialising/deserialising would add lots of
		817	overhead, as well as having to keep a proxy object everywhere.
		818
		819	Strings can easily be printed, easily serialised etc. and need no special
		820	procedures to be "valid".
		821
		822	And as a result, a miniport consists of a single closure stored in a
		823	global hash - it can't become much cheaper.
		824
		825	=item Why favour JSON, why not a real serialising format such as Storable?
		826
		827	In fact, any AnyEvent::MP node will happily accept Storable as framing
		828	format, but currently there is no way to make a node use Storable by
		829	default (although all nodes will accept it).
		830
		831	The default framing protocol is JSON because a) JSON::XS is many times
		832	faster for small messages and b) most importantly, after years of
		833	experience we found that object serialisation is causing more problems
		834	than it solves: Just like function calls, objects simply do not travel
		835	easily over the network, mostly because they will always be a copy, so you
		836	always have to re-think your design.
		837
		838	Keeping your messages simple, concentrating on data structures rather than
		839	objects, will keep your messages clean, tidy and efficient.
		840
		841	=back
		842
652	=head1 SEE ALSO	843	=head1 SEE ALSO
		844
		845	L<AnyEvent::MP::Intro> - a gentle introduction.
		846
		847	L<AnyEvent::MP::Kernel> - more, lower-level, stuff.
		848
		849	L<AnyEvent::MP::Global> - network maintainance and port groups, to find
		850	your applications.
653		851
654	L<AnyEvent>.	852	L<AnyEvent>.
655		853
656	=head1 AUTHOR	854	=head1 AUTHOR
657		855

Diff Legend

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

Comparing AnyEvent-MP/MP.pm (file contents): Revision 1.31 by root, Wed Aug 5 19:55:58 2009 UTC vs. Revision 1.73 by root, Mon Aug 31 11:08:25 2009 UTC

Diff Legend

Comparing AnyEvent-MP/MP.pm (file contents):
Revision 1.31 by root, Wed Aug 5 19:55:58 2009 UTC vs.
Revision 1.73 by root, Mon Aug 31 11:08:25 2009 UTC