ViewVC Help
View File | Revision Log | Show Annotations | Download File
/cvs/AnyEvent-MP/MP.pm
(Generate patch)

Comparing AnyEvent-MP/MP.pm (file contents):
Revision 1.22 by root, Tue Aug 4 18:33:30 2009 UTC vs.
Revision 1.49 by root, Thu Aug 13 15:29:58 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 initialise_node; # -OR-
17 initialise_node "localhost:4040"; # -OR-
18 initialise_node "slave/", "localhost:4040"
19
20 # ports are message endpoints
21
22 # sending messages
13 snd $port, type => data...; 23 snd $port, type => data...;
24 snd $port, @msg;
25 snd @msg_with_first_element_being_a_port;
14 26
15 $SELF # receiving/own port id in rcv callbacks 27 # creating/using miniports
28 my $miniport = port { my @msg = @_; 0 };
16 29
30 # creating/using full ports
31 my $port = port;
17 rcv $port, smartmatch => $cb->($port, @msg); 32 rcv $port, smartmatch => $cb->(@msg);
18
19 # examples:
20 rcv $port2, ping => sub { snd $_[0], "pong"; 0 }; 33 rcv $port, ping => sub { snd $_[0], "pong"; 0 };
21 rcv $port1, pong => sub { warn "pong received\n" }; 34 rcv $port, pong => sub { warn "pong received\n"; 0 };
22 snd $port2, ping => $port1;
23 35
24 # more, smarter, matches (_any_ is exported by this module) 36 # more, smarter, matches (_any_ is exported by this module)
25 rcv $port, [child_died => $pid] => sub { ... 37 rcv $port, [child_died => $pid] => sub { ...
26 rcv $port, [_any_, _any_, 3] => sub { .. $_[2] is 3 38 rcv $port, [_any_, _any_, 3] => sub { .. $_[2] is 3
27 39
40 # create a port on another node
41 my $port = spawn $node, $initfunc, @initdata;
42
43 # monitoring
44 mon $port, $cb->(@msg) # callback is invoked on death
45 mon $port, $otherport # kill otherport on abnormal death
46 mon $port, $otherport, @msg # send message on death
47
48=head1 CURRENT STATUS
49
50 AnyEvent::MP - stable API, should work
51 AnyEvent::MP::Intro - outdated
52 AnyEvent::MP::Kernel - WIP
53 AnyEvent::MP::Transport - mostly stable
54
55 stay tuned.
56
28=head1 DESCRIPTION 57=head1 DESCRIPTION
29 58
30This module (-family) implements a simple message passing framework. 59This module (-family) implements a simple message passing framework.
31 60
32Despite its simplicity, you can securely message other processes running 61Despite its simplicity, you can securely message other processes running
33on the same or other hosts. 62on the same or other hosts.
34 63
64For an introduction to this module family, see the L<AnyEvent::MP::Intro>
65manual page.
66
35At the moment, this module family is severly brokena nd underdocumented, 67At the moment, this module family is severly broken and underdocumented,
36so do not use. This was uploaded mainly to reserve the CPAN namespace - 68so do not use. This was uploaded mainly to reserve the CPAN namespace -
37stay tuned! 69stay tuned!
38 70
39=head1 CONCEPTS 71=head1 CONCEPTS
40 72
41=over 4 73=over 4
42 74
43=item port 75=item port
44 76
45A port is something you can send messages to with the C<snd> function, and 77A port is something you can send messages to (with the C<snd> function).
46you can register C<rcv> handlers with. All C<rcv> handlers will receive 78
47messages they match, messages will not be queued. 79Some ports allow you to register C<rcv> handlers that can match specific
80messages. All C<rcv> handlers will receive messages they match, messages
81will not be queued.
48 82
49=item port id - C<noderef#portname> 83=item port id - C<noderef#portname>
50 84
51A port id is always the noderef, a hash-mark (C<#>) as separator, followed 85A port id is normaly the concatenation of a noderef, a hash-mark (C<#>) as
52by a port name (a printable string of unspecified format). 86separator, and a port name (a printable string of unspecified format). An
87exception is the the node port, whose ID is identical to its node
88reference.
53 89
54=item node 90=item node
55 91
56A node is a single process containing at least one port - the node 92A node is a single process containing at least one port - the node
57port. You can send messages to node ports to let them create new ports, 93port. You can send messages to node ports to find existing ports or to
58among other things. 94create new ports, among other things.
59 95
60Initially, nodes are either private (single-process only) or hidden 96Nodes are either private (single-process only), slaves (connected to a
61(connected to a master node only). Only when they epxlicitly "become 97master node only) or public nodes (connectable from unrelated nodes).
62public" can you send them messages from unrelated other nodes.
63 98
64=item noderef - C<host:port,host:port...>, C<id@noderef>, C<id> 99=item noderef - C<host:port,host:port...>, C<id@noderef>, C<id>
65 100
66A noderef is a string that either uniquely identifies a given node (for 101A node reference is a string that either simply identifies the node (for
67private and hidden nodes), or contains a recipe on how to reach a given 102private and slave nodes), or contains a recipe on how to reach a given
68node (for public nodes). 103node (for public nodes).
69 104
105This recipe is simply a comma-separated list of C<address:port> pairs (for
106TCP/IP, other protocols might look different).
107
108Node references come in two flavours: resolved (containing only numerical
109addresses) or unresolved (where hostnames are used instead of addresses).
110
111Before using an unresolved node reference in a message you first have to
112resolve it.
113
70=back 114=back
71 115
72=head1 VARIABLES/FUNCTIONS 116=head1 VARIABLES/FUNCTIONS
73 117
74=over 4 118=over 4
75 119
76=cut 120=cut
77 121
78package AnyEvent::MP; 122package AnyEvent::MP;
79 123
80use AnyEvent::MP::Base; 124use AnyEvent::MP::Kernel;
81 125
82use common::sense; 126use common::sense;
83 127
84use Carp (); 128use Carp ();
85 129
86use AE (); 130use AE ();
87 131
88use base "Exporter"; 132use base "Exporter";
89 133
90our $VERSION = '0.02'; 134our $VERSION = $AnyEvent::MP::Kernel::VERSION;
135
91our @EXPORT = qw( 136our @EXPORT = qw(
92 NODE $NODE *SELF node_of _any_ 137 NODE $NODE *SELF node_of _any_
93 become_slave become_public 138 resolve_node initialise_node
94 snd rcv mon kil reg psub 139 snd rcv mon kil reg psub spawn
95 port 140 port
96); 141);
97 142
98our $SELF; 143our $SELF;
99 144
108The C<NODE> function returns, and the C<$NODE> variable contains 153The C<NODE> function returns, and the C<$NODE> variable contains
109the noderef of the local node. The value is initialised by a call 154the noderef of the local node. The value is initialised by a call
110to C<become_public> or C<become_slave>, after which all local port 155to C<become_public> or C<become_slave>, after which all local port
111identifiers become invalid. 156identifiers become invalid.
112 157
113=item $noderef = node_of $portid 158=item $noderef = node_of $port
114 159
115Extracts and returns the noderef from a portid or a noderef. 160Extracts and returns the noderef from a portid or a noderef.
161
162=item initialise_node $noderef, $seednode, $seednode...
163
164=item initialise_node "slave/", $master, $master...
165
166Before a node can talk to other nodes on the network it has to initialise
167itself - the minimum a node needs to know is it's own name, and optionally
168it should know the noderefs of some other nodes in the network.
169
170This function initialises a node - it must be called exactly once (or
171never) before calling other AnyEvent::MP functions.
172
173All arguments (optionally except for the first) are noderefs, which can be
174either resolved or unresolved.
175
176The first argument will be looked up in the configuration database first
177(if it is C<undef> then the current nodename will be used instead) to find
178the relevant configuration profile (see L<aemp>). If none is found then
179the default configuration is used. The configuration supplies additional
180seed/master nodes and can override the actual noderef.
181
182There are two types of networked nodes, public nodes and slave nodes:
183
184=over 4
185
186=item public nodes
187
188For public nodes, C<$noderef> (supplied either directly to
189C<initialise_node> or indirectly via a profile or the nodename) must be a
190noderef (possibly unresolved, in which case it will be resolved).
191
192After resolving, the node will bind itself on all endpoints and try to
193connect to all additional C<$seednodes> that are specified. Seednodes are
194optional and can be used to quickly bootstrap the node into an existing
195network.
196
197=item slave nodes
198
199When the C<$noderef> (either as given or overriden by the config file)
200is the special string C<slave/>, then the node will become a slave
201node. Slave nodes cannot be contacted from outside and will route most of
202their traffic to the master node that they attach to.
203
204At least one additional noderef is required (either by specifying it
205directly or because it is part of the configuration profile): The node
206will try to connect to all of them and will become a slave attached to the
207first node it can successfully connect to.
208
209=back
210
211This function will block until all nodes have been resolved and, for slave
212nodes, until it has successfully established a connection to a master
213server.
214
215Example: become a public node listening on the guessed noderef, or the one
216specified via C<aemp> for the current node. This should be the most common
217form of invocation for "daemon"-type nodes.
218
219 initialise_node;
220
221Example: become a slave node to any of the the seednodes specified via
222C<aemp>. This form is often used for commandline clients.
223
224 initialise_node "slave/";
225
226Example: become a slave node to any of the specified master servers. This
227form is also often used for commandline clients.
228
229 initialise_node "slave/", "master1", "192.168.13.17", "mp.example.net";
230
231Example: become a public node, and try to contact some well-known master
232servers to become part of the network.
233
234 initialise_node undef, "master1", "master2";
235
236Example: become a public node listening on port C<4041>.
237
238 initialise_node 4041;
239
240Example: become a public node, only visible on localhost port 4044.
241
242 initialise_node "localhost:4044";
243
244=item $cv = resolve_node $noderef
245
246Takes an unresolved node reference that may contain hostnames and
247abbreviated IDs, resolves all of them and returns a resolved node
248reference.
249
250In addition to C<address:port> pairs allowed in resolved noderefs, the
251following forms are supported:
252
253=over 4
254
255=item the empty string
256
257An empty-string component gets resolved as if the default port (4040) was
258specified.
259
260=item naked port numbers (e.g. C<1234>)
261
262These are resolved by prepending the local nodename and a colon, to be
263further resolved.
264
265=item hostnames (e.g. C<localhost:1234>, C<localhost>)
266
267These are resolved by using AnyEvent::DNS to resolve them, optionally
268looking up SRV records for the C<aemp=4040> port, if no port was
269specified.
270
271=back
116 272
117=item $SELF 273=item $SELF
118 274
119Contains the current port id while executing C<rcv> callbacks or C<psub> 275Contains the current port id while executing C<rcv> callbacks or C<psub>
120blocks. 276blocks.
123 279
124Due to some quirks in how perl exports variables, it is impossible to 280Due to some quirks in how perl exports variables, it is impossible to
125just export C<$SELF>, all the symbols called C<SELF> are exported by this 281just export C<$SELF>, all the symbols called C<SELF> are exported by this
126module, but only C<$SELF> is currently used. 282module, but only C<$SELF> is currently used.
127 283
128=item snd $portid, type => @data 284=item snd $port, type => @data
129 285
130=item snd $portid, @msg 286=item snd $port, @msg
131 287
132Send the given message to the given port ID, which can identify either 288Send the given message to the given port ID, which can identify either
133a local or a remote port, and can be either a string or soemthignt hat 289a local or a remote port, and can be either a string or soemthignt hat
134stringifies a sa port ID (such as a port object :). 290stringifies a sa port ID (such as a port object :).
135 291
145JSON is used, then only strings, numbers and arrays and hashes consisting 301JSON is used, then only strings, numbers and arrays and hashes consisting
146of those are allowed (no objects). When Storable is used, then anything 302of those are allowed (no objects). When Storable is used, then anything
147that Storable can serialise and deserialise is allowed, and for the local 303that Storable can serialise and deserialise is allowed, and for the local
148node, anything can be passed. 304node, anything can be passed.
149 305
150=item kil $portid[, @reason]
151
152Kill the specified port with the given C<@reason>.
153
154If no C<@reason> is specified, then the port is killed "normally" (linked
155ports will not be kileld, or even notified).
156
157Otherwise, linked ports get killed with the same reason (second form of
158C<mon>, see below).
159
160Runtime errors while evaluating C<rcv> callbacks or inside C<psub> blocks
161will be reported as reason C<< die => $@ >>.
162
163Transport/communication errors are reported as C<< transport_error =>
164$message >>.
165
166=item $guard = mon $portid, $cb->(@reason)
167
168=item $guard = mon $portid, $otherport
169
170=item $guard = mon $portid, $otherport, @msg
171
172Monitor the given port and do something when the port is killed.
173
174In the first form, the callback is simply called with any number
175of C<@reason> elements (no @reason means that the port was deleted
176"normally"). Note also that I<< the callback B<must> never die >>, so use
177C<eval> if unsure.
178
179In the second form, the other port will be C<kil>'ed with C<@reason>, iff
180a @reason was specified, i.e. on "normal" kils nothing happens, while
181under all other conditions, the other port is killed with the same reason.
182
183In the last form, a message of the form C<@msg, @reason> will be C<snd>.
184
185Example: call a given callback when C<$port> is killed.
186
187 mon $port, sub { warn "port died because of <@_>\n" };
188
189Example: kill ourselves when C<$port> is killed abnormally.
190
191 mon $port, $self;
192
193Example: send us a restart message another C<$port> is killed.
194
195 mon $port, $self => "restart";
196
197=cut
198
199sub mon {
200 my ($noderef, $port, $cb) = ((split /#/, shift, 2), shift);
201
202 my $node = $NODE{$noderef} || add_node $noderef;
203
204 #TODO: ports must not be references
205 if (!ref $cb or "AnyEvent::MP::Port" eq ref $cb) {
206 if (@_) {
207 # send a kill info message
208 my (@msg) = ($cb, @_);
209 $cb = sub { snd @msg, @_ };
210 } else {
211 # simply kill other port
212 my $port = $cb;
213 $cb = sub { kil $port, @_ if @_ };
214 }
215 }
216
217 $node->monitor ($port, $cb);
218
219 defined wantarray
220 and AnyEvent::Util::guard { $node->unmonitor ($port, $cb) }
221}
222
223=item $guard = mon_guard $port, $ref, $ref...
224
225Monitors the given C<$port> and keeps the passed references. When the port
226is killed, the references will be freed.
227
228Optionally returns a guard that will stop the monitoring.
229
230This function is useful when you create e.g. timers or other watchers and
231want to free them when the port gets killed:
232
233 $port->rcv (start => sub {
234 my $timer; $timer = mon_guard $port, AE::timer 1, 1, sub {
235 undef $timer if 0.9 < rand;
236 });
237 });
238
239=cut
240
241sub mon_guard {
242 my ($port, @refs) = @_;
243
244 mon $port, sub { 0 && @refs }
245}
246
247=item $local_port = port 306=item $local_port = port
248 307
249Create a new local port object that supports message matching. 308Create a new local port object that can be used either as a pattern
309matching port ("full port") or a single-callback port ("miniport"),
310depending on how C<rcv> callbacks are bound to the object.
250 311
251=item $portid = port { my @msg = @_; $finished } 312=item $port = port { my @msg = @_; $finished }
252 313
253Creates a "mini port", that is, a very lightweight port without any 314Creates a "miniport", that is, a very lightweight port without any pattern
254pattern matching behind it, and returns its ID. 315matching behind it, and returns its ID. Semantically the same as creating
316a port and calling C<rcv $port, $callback> on it.
255 317
256The block will be called for every message received on the port. When the 318The block will be called for every message received on the port. When the
257callback returns a true value its job is considered "done" and the port 319callback returns a true value its job is considered "done" and the port
258will be destroyed. Otherwise it will stay alive. 320will be destroyed. Otherwise it will stay alive.
259 321
260The message will be passed as-is, no extra argument (i.e. no port id) will 322The message will be passed as-is, no extra argument (i.e. no port id) will
261be passed to the callback. 323be passed to the callback.
262 324
263If you need the local port id in the callback, this works nicely: 325If you need the local port id in the callback, this works nicely:
264 326
265 my $port; $port = miniport { 327 my $port; $port = port {
266 snd $otherport, reply => $port; 328 snd $otherport, reply => $port;
267 }; 329 };
268 330
269=cut 331=cut
332
333sub rcv($@);
270 334
271sub port(;&) { 335sub port(;&) {
272 my $id = "$UNIQ." . $ID++; 336 my $id = "$UNIQ." . $ID++;
273 my $port = "$NODE#$id"; 337 my $port = "$NODE#$id";
274 338
275 if (@_) { 339 if (@_) {
340 rcv $port, shift;
341 } else {
342 $PORT{$id} = sub { }; # nop
343 }
344
345 $port
346}
347
348=item reg $port, $name
349
350=item reg $name
351
352Registers the given port (or C<$SELF><<< if missing) under the name
353C<$name>. If the name already exists it is replaced.
354
355A port can only be registered under one well known name.
356
357A port automatically becomes unregistered when it is killed.
358
359=cut
360
361sub reg(@) {
362 my $port = @_ > 1 ? shift : $SELF || Carp::croak 'reg: called with one argument only, but $SELF not set,';
363
364 $REG{$_[0]} = $port;
365}
366
367=item rcv $port, $callback->(@msg)
368
369Replaces the callback on the specified miniport (after converting it to
370one if required).
371
372=item rcv $port, tagstring => $callback->(@msg), ...
373
374=item rcv $port, $smartmatch => $callback->(@msg), ...
375
376=item rcv $port, [$smartmatch...] => $callback->(@msg), ...
377
378Register callbacks to be called on matching messages on the given full
379port (after converting it to one if required) and return the port.
380
381The callback has to return a true value when its work is done, after
382which is will be removed, or a false value in which case it will stay
383registered.
384
385The global C<$SELF> (exported by this module) contains C<$port> while
386executing the callback.
387
388Runtime errors during callback execution will result in the port being
389C<kil>ed.
390
391If the match is an array reference, then it will be matched against the
392first elements of the message, otherwise only the first element is being
393matched.
394
395Any element in the match that is specified as C<_any_> (a function
396exported by this module) matches any single element of the message.
397
398While not required, it is highly recommended that the first matching
399element is a string identifying the message. The one-string-only match is
400also the most efficient match (by far).
401
402Example: create a port and bind receivers on it in one go.
403
404 my $port = rcv port,
405 msg1 => sub { ...; 0 },
406 msg2 => sub { ...; 0 },
407 ;
408
409Example: create a port, bind receivers and send it in a message elsewhere
410in one go:
411
412 snd $otherport, reply =>
413 rcv port,
414 msg1 => sub { ...; 0 },
415 ...
416 ;
417
418=cut
419
420sub rcv($@) {
421 my $port = shift;
422 my ($noderef, $portid) = split /#/, $port, 2;
423
424 ($NODE{$noderef} || add_node $noderef) == $NODE{""}
425 or Carp::croak "$port: rcv can only be called on local ports, caught";
426
427 if (@_ == 1) {
276 my $cb = shift; 428 my $cb = shift;
429 delete $PORT_DATA{$portid};
277 $PORT{$id} = sub { 430 $PORT{$portid} = sub {
278 local $SELF = $port; 431 local $SELF = $port;
279 eval { 432 eval {
280 &$cb 433 &$cb
281 and kil $id; 434 and kil $port;
282 }; 435 };
283 _self_die if $@; 436 _self_die if $@;
284 }; 437 };
285 } else { 438 } else {
439 my $self = $PORT_DATA{$portid} ||= do {
286 my $self = bless { 440 my $self = bless {
287 id => "$NODE#$id", 441 id => $port,
288 }, "AnyEvent::MP::Port"; 442 }, "AnyEvent::MP::Port";
289 443
290 $PORT_DATA{$id} = $self;
291 $PORT{$id} = sub { 444 $PORT{$portid} = sub {
292 local $SELF = $port; 445 local $SELF = $port;
293 446
294 eval { 447 eval {
295 for (@{ $self->{rc0}{$_[0]} }) { 448 for (@{ $self->{rc0}{$_[0]} }) {
296 $_ && &{$_->[0]} 449 $_ && &{$_->[0]}
297 && undef $_; 450 && undef $_;
298 } 451 }
299 452
300 for (@{ $self->{rcv}{$_[0]} }) { 453 for (@{ $self->{rcv}{$_[0]} }) {
301 $_ && [@_[1 .. @{$_->[1]}]] ~~ $_->[1] 454 $_ && [@_[1 .. @{$_->[1]}]] ~~ $_->[1]
302 && &{$_->[0]} 455 && &{$_->[0]}
303 && undef $_; 456 && undef $_;
304 } 457 }
305 458
306 for (@{ $self->{any} }) { 459 for (@{ $self->{any} }) {
307 $_ && [@_[0 .. $#{$_->[1]}]] ~~ $_->[1] 460 $_ && [@_[0 .. $#{$_->[1]}]] ~~ $_->[1]
308 && &{$_->[0]} 461 && &{$_->[0]}
309 && undef $_; 462 && undef $_;
463 }
310 } 464 };
465 _self_die if $@;
311 }; 466 };
312 _self_die if $@; 467
468 $self
313 }; 469 };
314 }
315 470
316 $port
317}
318
319=item reg $portid, $name
320
321Registers the given port under the name C<$name>. If the name already
322exists it is replaced.
323
324A port can only be registered under one well known name.
325
326A port automatically becomes unregistered when it is killed.
327
328=cut
329
330sub reg(@) {
331 my ($portid, $name) = @_;
332
333 $REG{$name} = $portid;
334}
335
336=item rcv $portid, tagstring => $callback->(@msg), ...
337
338=item rcv $portid, $smartmatch => $callback->(@msg), ...
339
340=item rcv $portid, [$smartmatch...] => $callback->(@msg), ...
341
342Register callbacks to be called on matching messages on the given port.
343
344The callback has to return a true value when its work is done, after
345which is will be removed, or a false value in which case it will stay
346registered.
347
348The global C<$SELF> (exported by this module) contains C<$portid> while
349executing the callback.
350
351Runtime errors wdurign callback execution will result in the port being
352C<kil>ed.
353
354If the match is an array reference, then it will be matched against the
355first elements of the message, otherwise only the first element is being
356matched.
357
358Any element in the match that is specified as C<_any_> (a function
359exported by this module) matches any single element of the message.
360
361While not required, it is highly recommended that the first matching
362element is a string identifying the message. The one-string-only match is
363also the most efficient match (by far).
364
365=cut
366
367sub rcv($@) {
368 my ($noderef, $port) = split /#/, shift, 2;
369
370 ($NODE{$noderef} || add_node $noderef) == $NODE{""}
371 or Carp::croak "$noderef#$port: rcv can only be called on local ports, caught";
372
373 my $self = $PORT_DATA{$port}
374 or Carp::croak "$noderef#$port: rcv can only be called on message matching ports, caught";
375
376 "AnyEvent::MP::Port" eq ref $self 471 "AnyEvent::MP::Port" eq ref $self
377 or Carp::croak "$noderef#$port: rcv can only be called on message matching ports, caught"; 472 or Carp::croak "$port: rcv can only be called on message matching ports, caught";
378 473
379 while (@_) { 474 while (@_) {
380 my ($match, $cb) = splice @_, 0, 2; 475 my ($match, $cb) = splice @_, 0, 2;
381 476
382 if (!ref $match) { 477 if (!ref $match) {
383 push @{ $self->{rc0}{$match} }, [$cb]; 478 push @{ $self->{rc0}{$match} }, [$cb];
384 } elsif (("ARRAY" eq ref $match && !ref $match->[0])) { 479 } elsif (("ARRAY" eq ref $match && !ref $match->[0])) {
385 my ($type, @match) = @$match; 480 my ($type, @match) = @$match;
386 @match 481 @match
387 ? push @{ $self->{rcv}{$match->[0]} }, [$cb, \@match] 482 ? push @{ $self->{rcv}{$match->[0]} }, [$cb, \@match]
388 : push @{ $self->{rc0}{$match->[0]} }, [$cb]; 483 : push @{ $self->{rc0}{$match->[0]} }, [$cb];
389 } else { 484 } else {
390 push @{ $self->{any} }, [$cb, $match]; 485 push @{ $self->{any} }, [$cb, $match];
486 }
391 } 487 }
392 } 488 }
489
490 $port
393} 491}
394 492
395=item $closure = psub { BLOCK } 493=item $closure = psub { BLOCK }
396 494
397Remembers C<$SELF> and creates a closure out of the BLOCK. When the 495Remembers C<$SELF> and creates a closure out of the BLOCK. When the
428 $res 526 $res
429 } 527 }
430 } 528 }
431} 529}
432 530
433=back 531=item $guard = mon $port, $cb->(@reason)
434 532
435=head1 FUNCTIONS FOR NODES 533=item $guard = mon $port, $rcvport
436 534
437=over 4 535=item $guard = mon $port
438 536
439=item become_public endpoint... 537=item $guard = mon $port, $rcvport, @msg
440 538
441Tells the node to become a public node, i.e. reachable from other nodes. 539Monitor the given port and do something when the port is killed or
540messages to it were lost, and optionally return a guard that can be used
541to stop monitoring again.
442 542
443If no arguments are given, or the first argument is C<undef>, then 543C<mon> effectively guarantees that, in the absence of hardware failures,
444AnyEvent::MP tries to bind on port C<4040> on all IP addresses that the 544that after starting the monitor, either all messages sent to the port
445local nodename resolves to. 545will arrive, or the monitoring action will be invoked after possible
546message loss has been detected. No messages will be lost "in between"
547(after the first lost message no further messages will be received by the
548port). After the monitoring action was invoked, further messages might get
549delivered again.
446 550
447Otherwise the first argument must be an array-reference with transport 551In the first form (callback), the callback is simply called with any
448endpoints ("ip:port", "hostname:port") or port numbers (in which case the 552number of C<@reason> elements (no @reason means that the port was deleted
449local nodename is used as hostname). The endpoints are all resolved and 553"normally"). Note also that I<< the callback B<must> never die >>, so use
450will become the node reference. 554C<eval> if unsure.
451 555
556In the second form (another port given), the other port (C<$rcvport>)
557will be C<kil>'ed with C<@reason>, iff a @reason was specified, i.e. on
558"normal" kils nothing happens, while under all other conditions, the other
559port is killed with the same reason.
560
561The third form (kill self) is the same as the second form, except that
562C<$rvport> defaults to C<$SELF>.
563
564In the last form (message), a message of the form C<@msg, @reason> will be
565C<snd>.
566
567As a rule of thumb, monitoring requests should always monitor a port from
568a local port (or callback). The reason is that kill messages might get
569lost, just like any other message. Another less obvious reason is that
570even monitoring requests can get lost (for exmaple, when the connection
571to the other node goes down permanently). When monitoring a port locally
572these problems do not exist.
573
574Example: call a given callback when C<$port> is killed.
575
576 mon $port, sub { warn "port died because of <@_>\n" };
577
578Example: kill ourselves when C<$port> is killed abnormally.
579
580 mon $port;
581
582Example: send us a restart message when another C<$port> is killed.
583
584 mon $port, $self => "restart";
585
452=cut 586=cut
587
588sub mon {
589 my ($noderef, $port) = split /#/, shift, 2;
590
591 my $node = $NODE{$noderef} || add_node $noderef;
592
593 my $cb = @_ ? shift : $SELF || Carp::croak 'mon: called with one argument only, but $SELF not set,';
594
595 unless (ref $cb) {
596 if (@_) {
597 # send a kill info message
598 my (@msg) = ($cb, @_);
599 $cb = sub { snd @msg, @_ };
600 } else {
601 # simply kill other port
602 my $port = $cb;
603 $cb = sub { kil $port, @_ if @_ };
604 }
605 }
606
607 $node->monitor ($port, $cb);
608
609 defined wantarray
610 and AnyEvent::Util::guard { $node->unmonitor ($port, $cb) }
611}
612
613=item $guard = mon_guard $port, $ref, $ref...
614
615Monitors the given C<$port> and keeps the passed references. When the port
616is killed, the references will be freed.
617
618Optionally returns a guard that will stop the monitoring.
619
620This function is useful when you create e.g. timers or other watchers and
621want to free them when the port gets killed:
622
623 $port->rcv (start => sub {
624 my $timer; $timer = mon_guard $port, AE::timer 1, 1, sub {
625 undef $timer if 0.9 < rand;
626 });
627 });
628
629=cut
630
631sub mon_guard {
632 my ($port, @refs) = @_;
633
634 #TODO: mon-less form?
635
636 mon $port, sub { 0 && @refs }
637}
638
639=item kil $port[, @reason]
640
641Kill the specified port with the given C<@reason>.
642
643If no C<@reason> is specified, then the port is killed "normally" (linked
644ports will not be kileld, or even notified).
645
646Otherwise, linked ports get killed with the same reason (second form of
647C<mon>, see below).
648
649Runtime errors while evaluating C<rcv> callbacks or inside C<psub> blocks
650will be reported as reason C<< die => $@ >>.
651
652Transport/communication errors are reported as C<< transport_error =>
653$message >>.
654
655=cut
656
657=item $port = spawn $node, $initfunc[, @initdata]
658
659Creates a port on the node C<$node> (which can also be a port ID, in which
660case it's the node where that port resides).
661
662The port ID of the newly created port is return immediately, and it is
663permissible to immediately start sending messages or monitor the port.
664
665After the port has been created, the init function is
666called. This function must be a fully-qualified function name
667(e.g. C<MyApp::Chat::Server::init>). To specify a function in the main
668program, use C<::name>.
669
670If the function doesn't exist, then the node tries to C<require>
671the package, then the package above the package and so on (e.g.
672C<MyApp::Chat::Server>, C<MyApp::Chat>, C<MyApp>) until the function
673exists or it runs out of package names.
674
675The init function is then called with the newly-created port as context
676object (C<$SELF>) and the C<@initdata> values as arguments.
677
678A common idiom is to pass your own port, monitor the spawned port, and
679in the init function, monitor the original port. This two-way monitoring
680ensures that both ports get cleaned up when there is a problem.
681
682Example: spawn a chat server port on C<$othernode>.
683
684 # this node, executed from within a port context:
685 my $server = spawn $othernode, "MyApp::Chat::Server::connect", $SELF;
686 mon $server;
687
688 # init function on C<$othernode>
689 sub connect {
690 my ($srcport) = @_;
691
692 mon $srcport;
693
694 rcv $SELF, sub {
695 ...
696 };
697 }
698
699=cut
700
701sub _spawn {
702 my $port = shift;
703 my $init = shift;
704
705 local $SELF = "$NODE#$port";
706 eval {
707 &{ load_func $init }
708 };
709 _self_die if $@;
710}
711
712sub spawn(@) {
713 my ($noderef, undef) = split /#/, shift, 2;
714
715 my $id = "$RUNIQ." . $ID++;
716
717 $_[0] =~ /::/
718 or Carp::croak "spawn init function must be a fully-qualified name, caught";
719
720 ($NODE{$noderef} || add_node $noderef)
721 ->send (["", "AnyEvent::MP::_spawn" => $id, @_]);
722
723 "$noderef#$id"
724}
453 725
454=back 726=back
455 727
456=head1 NODE MESSAGES 728=head1 NODE MESSAGES
457 729
458Nodes understand the following messages sent to them. Many of them take 730Nodes understand the following messages sent to them. Many of them take
459arguments called C<@reply>, which will simply be used to compose a reply 731arguments called C<@reply>, which will simply be used to compose a reply
460message - C<$reply[0]> is the port to reply to, C<$reply[1]> the type and 732message - C<$reply[0]> is the port to reply to, C<$reply[1]> the type and
461the remaining arguments are simply the message data. 733the remaining arguments are simply the message data.
462 734
735While other messages exist, they are not public and subject to change.
736
463=over 4 737=over 4
464 738
465=cut 739=cut
466 740
467=item lookup => $name, @reply 741=item lookup => $name, @reply
495 snd $NODE, time => $myport, timereply => 1, 2; 769 snd $NODE, time => $myport, timereply => 1, 2;
496 # => snd $myport, timereply => 1, 2, <time> 770 # => snd $myport, timereply => 1, 2, <time>
497 771
498=back 772=back
499 773
774=head1 AnyEvent::MP vs. Distributed Erlang
775
776AnyEvent::MP got lots of its ideas from distributed Erlang (Erlang node
777== aemp node, Erlang process == aemp port), so many of the documents and
778programming techniques employed by Erlang apply to AnyEvent::MP. Here is a
779sample:
780
781 http://www.Erlang.se/doc/programming_rules.shtml
782 http://Erlang.org/doc/getting_started/part_frame.html # chapters 3 and 4
783 http://Erlang.org/download/Erlang-book-part1.pdf # chapters 5 and 6
784 http://Erlang.org/download/armstrong_thesis_2003.pdf # chapters 4 and 5
785
786Despite the similarities, there are also some important differences:
787
788=over 4
789
790=item * Node references contain the recipe on how to contact them.
791
792Erlang relies on special naming and DNS to work everywhere in the
793same way. AEMP relies on each node knowing it's own address(es), with
794convenience functionality.
795
796This means that AEMP requires a less tightly controlled environment at the
797cost of longer node references and a slightly higher management overhead.
798
799=item * Erlang uses processes and a mailbox, AEMP does not queue.
800
801Erlang uses processes that selctively receive messages, and therefore
802needs a queue. AEMP is event based, queuing messages would serve no useful
803purpose.
804
805(But see L<Coro::MP> for a more Erlang-like process model on top of AEMP).
806
807=item * Erlang sends are synchronous, AEMP sends are asynchronous.
808
809Sending messages in Erlang is synchronous and blocks the process. AEMP
810sends are immediate, connection establishment is handled in the
811background.
812
813=item * Erlang can silently lose messages, AEMP cannot.
814
815Erlang makes few guarantees on messages delivery - messages can get lost
816without any of the processes realising it (i.e. you send messages a, b,
817and c, and the other side only receives messages a and c).
818
819AEMP guarantees correct ordering, and the guarantee that there are no
820holes in the message sequence.
821
822=item * In Erlang, processes can be declared dead and later be found to be
823alive.
824
825In Erlang it can happen that a monitored process is declared dead and
826linked processes get killed, but later it turns out that the process is
827still alive - and can receive messages.
828
829In AEMP, when port monitoring detects a port as dead, then that port will
830eventually be killed - it cannot happen that a node detects a port as dead
831and then later sends messages to it, finding it is still alive.
832
833=item * Erlang can send messages to the wrong port, AEMP does not.
834
835In Erlang it is quite possible that a node that restarts reuses a process
836ID known to other nodes for a completely different process, causing
837messages destined for that process to end up in an unrelated process.
838
839AEMP never reuses port IDs, so old messages or old port IDs floating
840around in the network will not be sent to an unrelated port.
841
842=item * Erlang uses unprotected connections, AEMP uses secure
843authentication and can use TLS.
844
845AEMP can use a proven protocol - SSL/TLS - to protect connections and
846securely authenticate nodes.
847
848=item * The AEMP protocol is optimised for both text-based and binary
849communications.
850
851The AEMP protocol, unlike the Erlang protocol, supports both
852language-independent text-only protocols (good for debugging) and binary,
853language-specific serialisers (e.g. Storable).
854
855It has also been carefully designed to be implementable in other languages
856with a minimum of work while gracefully degrading fucntionality to make the
857protocol simple.
858
859=item * AEMP has more flexible monitoring options than Erlang.
860
861In Erlang, you can chose to receive I<all> exit signals as messages
862or I<none>, there is no in-between, so monitoring single processes is
863difficult to implement. Monitoring in AEMP is more flexible than in
864Erlang, as one can choose between automatic kill, exit message or callback
865on a per-process basis.
866
867=item * Erlang tries to hide remote/local connections, AEMP does not.
868
869Monitoring in Erlang is not an indicator of process death/crashes,
870as linking is (except linking is unreliable in Erlang).
871
872In AEMP, you don't "look up" registered port names or send to named ports
873that might or might not be persistent. Instead, you normally spawn a port
874on the remote node. The init function monitors the you, and you monitor
875the remote port. Since both monitors are local to the node, they are much
876more reliable.
877
878This also saves round-trips and avoids sending messages to the wrong port
879(hard to do in Erlang).
880
881=back
882
883=head1 RATIONALE
884
885=over 4
886
887=item Why strings for ports and noderefs, why not objects?
888
889We considered "objects", but found that the actual number of methods
890thatc an be called are very low. Since port IDs and noderefs travel over
891the network frequently, the serialising/deserialising would add lots of
892overhead, as well as having to keep a proxy object.
893
894Strings can easily be printed, easily serialised etc. and need no special
895procedures to be "valid".
896
897And a a miniport consists of a single closure stored in a global hash - it
898can't become much cheaper.
899
900=item Why favour JSON, why not real serialising format such as Storable?
901
902In fact, any AnyEvent::MP node will happily accept Storable as framing
903format, but currently there is no way to make a node use Storable by
904default.
905
906The default framing protocol is JSON because a) JSON::XS is many times
907faster for small messages and b) most importantly, after years of
908experience we found that object serialisation is causing more problems
909than it gains: Just like function calls, objects simply do not travel
910easily over the network, mostly because they will always be a copy, so you
911always have to re-think your design.
912
913Keeping your messages simple, concentrating on data structures rather than
914objects, will keep your messages clean, tidy and efficient.
915
916=back
917
500=head1 SEE ALSO 918=head1 SEE ALSO
501 919
502L<AnyEvent>. 920L<AnyEvent>.
503 921
504=head1 AUTHOR 922=head1 AUTHOR

Diff Legend

Removed lines
+ Added lines
< Changed lines
> Changed lines