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

Comparing AnyEvent/lib/AnyEvent/Handle.pm (file contents):
Revision 1.27 by root, Sat May 24 15:26:04 2008 UTC vs.
Revision 1.102 by root, Wed Oct 29 14:32:02 2008 UTC

1package AnyEvent::Handle; 1package AnyEvent::Handle;
2 2
3no warnings; 3no warnings;
4use strict; 4use strict qw(subs vars);
5 5
6use AnyEvent (); 6use AnyEvent ();
7use AnyEvent::Util (); 7use AnyEvent::Util qw(WSAEWOULDBLOCK);
8use Scalar::Util (); 8use Scalar::Util ();
9use Carp (); 9use Carp ();
10use Fcntl (); 10use Fcntl ();
11use Errno qw/EAGAIN EINTR/; 11use Errno qw(EAGAIN EINTR);
12 12
13=head1 NAME 13=head1 NAME
14 14
15AnyEvent::Handle - non-blocking I/O on file handles via AnyEvent 15AnyEvent::Handle - non-blocking I/O on file handles via AnyEvent
16 16
17This module is experimental.
18
19=cut 17=cut
20 18
21our $VERSION = '0.04'; 19our $VERSION = 4.3;
22 20
23=head1 SYNOPSIS 21=head1 SYNOPSIS
24 22
25 use AnyEvent; 23 use AnyEvent;
26 use AnyEvent::Handle; 24 use AnyEvent::Handle;
27 25
28 my $cv = AnyEvent->condvar; 26 my $cv = AnyEvent->condvar;
29 27
30 my $ae_fh = AnyEvent::Handle->new (fh => \*STDIN); 28 my $handle =
31
32 #TODO
33
34 # or use the constructor to pass the callback:
35
36 my $ae_fh2 =
37 AnyEvent::Handle->new ( 29 AnyEvent::Handle->new (
38 fh => \*STDIN, 30 fh => \*STDIN,
39 on_eof => sub { 31 on_eof => sub {
40 $cv->broadcast; 32 $cv->broadcast;
41 }, 33 },
42 #TODO
43 ); 34 );
44 35
45 $cv->wait; 36 # send some request line
37 $handle->push_write ("getinfo\015\012");
38
39 # read the response line
40 $handle->push_read (line => sub {
41 my ($handle, $line) = @_;
42 warn "read line <$line>\n";
43 $cv->send;
44 });
45
46 $cv->recv;
46 47
47=head1 DESCRIPTION 48=head1 DESCRIPTION
48 49
49This module is a helper module to make it easier to do event-based I/O on 50This module is a helper module to make it easier to do event-based I/O on
50filehandles. For utility functions for doing non-blocking connects and accepts 51filehandles. For utility functions for doing non-blocking connects and accepts
51on sockets see L<AnyEvent::Util>. 52on sockets see L<AnyEvent::Util>.
52 53
54The L<AnyEvent::Intro> tutorial contains some well-documented
55AnyEvent::Handle examples.
56
53In the following, when the documentation refers to of "bytes" then this 57In the following, when the documentation refers to of "bytes" then this
54means characters. As sysread and syswrite are used for all I/O, their 58means characters. As sysread and syswrite are used for all I/O, their
55treatment of characters applies to this module as well. 59treatment of characters applies to this module as well.
56 60
57All callbacks will be invoked with the handle object as their first 61All callbacks will be invoked with the handle object as their first
69 73
70=item fh => $filehandle [MANDATORY] 74=item fh => $filehandle [MANDATORY]
71 75
72The filehandle this L<AnyEvent::Handle> object will operate on. 76The filehandle this L<AnyEvent::Handle> object will operate on.
73 77
74NOTE: The filehandle will be set to non-blocking (using 78NOTE: The filehandle will be set to non-blocking mode (using
75AnyEvent::Util::fh_nonblocking). 79C<AnyEvent::Util::fh_nonblocking>) by the constructor and needs to stay in
80that mode.
76 81
77=item on_eof => $cb->($self) 82=item on_eof => $cb->($handle)
78 83
79Set the callback to be called on EOF. 84Set the callback to be called when an end-of-file condition is detected,
85i.e. in the case of a socket, when the other side has closed the
86connection cleanly.
80 87
88For sockets, this just means that the other side has stopped sending data,
89you can still try to write data, and, in fact, one can return from the EOF
90callback and continue writing data, as only the read part has been shut
91down.
92
81While not mandatory, it is highly recommended to set an eof callback, 93While not mandatory, it is I<highly> recommended to set an EOF callback,
82otherwise you might end up with a closed socket while you are still 94otherwise you might end up with a closed socket while you are still
83waiting for data. 95waiting for data.
84 96
97If an EOF condition has been detected but no C<on_eof> callback has been
98set, then a fatal error will be raised with C<$!> set to <0>.
99
85=item on_error => $cb->($self) 100=item on_error => $cb->($handle, $fatal)
86 101
87This is the fatal error callback, that is called when, well, a fatal error 102This is the error callback, which is called when, well, some error
88occurs, such as not being able to resolve the hostname, failure to connect 103occured, such as not being able to resolve the hostname, failure to
89or a read error. 104connect or a read error.
90 105
91The object will not be in a usable state when this callback has been 106Some errors are fatal (which is indicated by C<$fatal> being true). On
92called. 107fatal errors the handle object will be shut down and will not be usable
108(but you are free to look at the current C<< ->rbuf >>). Examples of fatal
109errors are an EOF condition with active (but unsatisifable) read watchers
110(C<EPIPE>) or I/O errors.
111
112Non-fatal errors can be retried by simply returning, but it is recommended
113to simply ignore this parameter and instead abondon the handle object
114when this callback is invoked. Examples of non-fatal errors are timeouts
115C<ETIMEDOUT>) or badly-formatted data (C<EBADMSG>).
93 116
94On callback entrance, the value of C<$!> contains the operating system 117On callback entrance, the value of C<$!> contains the operating system
95error (or C<ENOSPC> or C<EPIPE>). 118error (or C<ENOSPC>, C<EPIPE>, C<ETIMEDOUT> or C<EBADMSG>).
96 119
97While not mandatory, it is I<highly> recommended to set this callback, as 120While not mandatory, it is I<highly> recommended to set this callback, as
98you will not be notified of errors otherwise. The default simply calls 121you will not be notified of errors otherwise. The default simply calls
99die. 122C<croak>.
100 123
101=item on_read => $cb->($self) 124=item on_read => $cb->($handle)
102 125
103This sets the default read callback, which is called when data arrives 126This sets the default read callback, which is called when data arrives
104and no read request is in the queue. 127and no read request is in the queue (unlike read queue callbacks, this
128callback will only be called when at least one octet of data is in the
129read buffer).
105 130
106To access (and remove data from) the read buffer, use the C<< ->rbuf >> 131To access (and remove data from) the read buffer, use the C<< ->rbuf >>
107method or access the C<$self->{rbuf}> member directly. 132method or access the C<$handle->{rbuf}> member directly.
108 133
109When an EOF condition is detected then AnyEvent::Handle will first try to 134When an EOF condition is detected then AnyEvent::Handle will first try to
110feed all the remaining data to the queued callbacks and C<on_read> before 135feed all the remaining data to the queued callbacks and C<on_read> before
111calling the C<on_eof> callback. If no progress can be made, then a fatal 136calling the C<on_eof> callback. If no progress can be made, then a fatal
112error will be raised (with C<$!> set to C<EPIPE>). 137error will be raised (with C<$!> set to C<EPIPE>).
113 138
114=item on_drain => $cb->() 139=item on_drain => $cb->($handle)
115 140
116This sets the callback that is called when the write buffer becomes empty 141This sets the callback that is called when the write buffer becomes empty
117(or when the callback is set and the buffer is empty already). 142(or when the callback is set and the buffer is empty already).
118 143
119To append to the write buffer, use the C<< ->push_write >> method. 144To append to the write buffer, use the C<< ->push_write >> method.
120 145
146This callback is useful when you don't want to put all of your write data
147into the queue at once, for example, when you want to write the contents
148of some file to the socket you might not want to read the whole file into
149memory and push it into the queue, but instead only read more data from
150the file when the write queue becomes empty.
151
152=item timeout => $fractional_seconds
153
154If non-zero, then this enables an "inactivity" timeout: whenever this many
155seconds pass without a successful read or write on the underlying file
156handle, the C<on_timeout> callback will be invoked (and if that one is
157missing, a non-fatal C<ETIMEDOUT> error will be raised).
158
159Note that timeout processing is also active when you currently do not have
160any outstanding read or write requests: If you plan to keep the connection
161idle then you should disable the timout temporarily or ignore the timeout
162in the C<on_timeout> callback, in which case AnyEvent::Handle will simply
163restart the timeout.
164
165Zero (the default) disables this timeout.
166
167=item on_timeout => $cb->($handle)
168
169Called whenever the inactivity timeout passes. If you return from this
170callback, then the timeout will be reset as if some activity had happened,
171so this condition is not fatal in any way.
172
121=item rbuf_max => <bytes> 173=item rbuf_max => <bytes>
122 174
123If defined, then a fatal error will be raised (with C<$!> set to C<ENOSPC>) 175If defined, then a fatal error will be raised (with C<$!> set to C<ENOSPC>)
124when the read buffer ever (strictly) exceeds this size. This is useful to 176when the read buffer ever (strictly) exceeds this size. This is useful to
125avoid denial-of-service attacks. 177avoid some forms of denial-of-service attacks.
126 178
127For example, a server accepting connections from untrusted sources should 179For example, a server accepting connections from untrusted sources should
128be configured to accept only so-and-so much data that it cannot act on 180be configured to accept only so-and-so much data that it cannot act on
129(for example, when expecting a line, an attacker could send an unlimited 181(for example, when expecting a line, an attacker could send an unlimited
130amount of data without a callback ever being called as long as the line 182amount of data without a callback ever being called as long as the line
131isn't finished). 183isn't finished).
132 184
185=item autocork => <boolean>
186
187When disabled (the default), then C<push_write> will try to immediately
188write the data to the handle, if possible. This avoids having to register
189a write watcher and wait for the next event loop iteration, but can
190be inefficient if you write multiple small chunks (on the wire, this
191disadvantage is usually avoided by your kernel's nagle algorithm, see
192C<no_delay>, but this option can save costly syscalls).
193
194When enabled, then writes will always be queued till the next event loop
195iteration. This is efficient when you do many small writes per iteration,
196but less efficient when you do a single write only per iteration (or when
197the write buffer often is full). It also increases write latency.
198
199=item no_delay => <boolean>
200
201When doing small writes on sockets, your operating system kernel might
202wait a bit for more data before actually sending it out. This is called
203the Nagle algorithm, and usually it is beneficial.
204
205In some situations you want as low a delay as possible, which can be
206accomplishd by setting this option to a true value.
207
208The default is your opertaing system's default behaviour (most likely
209enabled), this option explicitly enables or disables it, if possible.
210
133=item read_size => <bytes> 211=item read_size => <bytes>
134 212
135The default read block size (the amount of bytes this module will try to read 213The default read block size (the amount of bytes this module will
136on each [loop iteration). Default: C<4096>. 214try to read during each loop iteration, which affects memory
215requirements). Default: C<8192>.
137 216
138=item low_water_mark => <bytes> 217=item low_water_mark => <bytes>
139 218
140Sets the amount of bytes (default: C<0>) that make up an "empty" write 219Sets the amount of bytes (default: C<0>) that make up an "empty" write
141buffer: If the write reaches this size or gets even samller it is 220buffer: If the write reaches this size or gets even samller it is
142considered empty. 221considered empty.
143 222
223Sometimes it can be beneficial (for performance reasons) to add data to
224the write buffer before it is fully drained, but this is a rare case, as
225the operating system kernel usually buffers data as well, so the default
226is good in almost all cases.
227
228=item linger => <seconds>
229
230If non-zero (default: C<3600>), then the destructor of the
231AnyEvent::Handle object will check whether there is still outstanding
232write data and will install a watcher that will write this data to the
233socket. No errors will be reported (this mostly matches how the operating
234system treats outstanding data at socket close time).
235
236This will not work for partial TLS data that could not be encoded
237yet. This data will be lost. Calling the C<stoptls> method in time might
238help.
239
144=item tls => "accept" | "connect" | Net::SSLeay::SSL object 240=item tls => "accept" | "connect" | Net::SSLeay::SSL object
145 241
146When this parameter is given, it enables TLS (SSL) mode, that means it 242When this parameter is given, it enables TLS (SSL) mode, that means
147will start making tls handshake and will transparently encrypt/decrypt 243AnyEvent will start a TLS handshake as soon as the conenction has been
148data. 244established and will transparently encrypt/decrypt data afterwards.
149 245
150TLS mode requires Net::SSLeay to be installed (it will be loaded 246TLS mode requires Net::SSLeay to be installed (it will be loaded
151automatically when you try to create a TLS handle). 247automatically when you try to create a TLS handle): this module doesn't
248have a dependency on that module, so if your module requires it, you have
249to add the dependency yourself.
152 250
153For the TLS server side, use C<accept>, and for the TLS client side of a 251Unlike TCP, TLS has a server and client side: for the TLS server side, use
154connection, use C<connect> mode. 252C<accept>, and for the TLS client side of a connection, use C<connect>
253mode.
155 254
156You can also provide your own TLS connection object, but you have 255You can also provide your own TLS connection object, but you have
157to make sure that you call either C<Net::SSLeay::set_connect_state> 256to make sure that you call either C<Net::SSLeay::set_connect_state>
158or C<Net::SSLeay::set_accept_state> on it before you pass it to 257or C<Net::SSLeay::set_accept_state> on it before you pass it to
159AnyEvent::Handle. 258AnyEvent::Handle.
160 259
161See the C<starttls> method if you need to start TLs negotiation later. 260See the C<< ->starttls >> method for when need to start TLS negotiation later.
162 261
163=item tls_ctx => $ssl_ctx 262=item tls_ctx => $ssl_ctx
164 263
165Use the given Net::SSLeay::CTX object to create the new TLS connection 264Use the given C<Net::SSLeay::CTX> object to create the new TLS connection
166(unless a connection object was specified directly). If this parameter is 265(unless a connection object was specified directly). If this parameter is
167missing, then AnyEvent::Handle will use C<AnyEvent::Handle::TLS_CTX>. 266missing, then AnyEvent::Handle will use C<AnyEvent::Handle::TLS_CTX>.
168 267
268=item json => JSON or JSON::XS object
269
270This is the json coder object used by the C<json> read and write types.
271
272If you don't supply it, then AnyEvent::Handle will create and use a
273suitable one (on demand), which will write and expect UTF-8 encoded JSON
274texts.
275
276Note that you are responsible to depend on the JSON module if you want to
277use this functionality, as AnyEvent does not have a dependency itself.
278
169=back 279=back
170 280
171=cut 281=cut
172 282
173sub new { 283sub new {
177 287
178 $self->{fh} or Carp::croak "mandatory argument fh is missing"; 288 $self->{fh} or Carp::croak "mandatory argument fh is missing";
179 289
180 AnyEvent::Util::fh_nonblocking $self->{fh}, 1; 290 AnyEvent::Util::fh_nonblocking $self->{fh}, 1;
181 291
182 if ($self->{tls}) {
183 require Net::SSLeay;
184 $self->starttls (delete $self->{tls}, delete $self->{tls_ctx}); 292 $self->starttls (delete $self->{tls}, delete $self->{tls_ctx})
185 } 293 if $self->{tls};
186 294
187 $self->on_eof (delete $self->{on_eof} ) if $self->{on_eof}; 295 $self->{_activity} = AnyEvent->now;
188 $self->on_error (delete $self->{on_error}) if $self->{on_error}; 296 $self->_timeout;
297
189 $self->on_drain (delete $self->{on_drain}) if $self->{on_drain}; 298 $self->on_drain (delete $self->{on_drain}) if exists $self->{on_drain};
190 $self->on_read (delete $self->{on_read} ) if $self->{on_read}; 299 $self->no_delay (delete $self->{no_delay}) if exists $self->{no_delay};
191 300
192 $self->start_read; 301 $self->start_read
302 if $self->{on_read};
193 303
194 $self 304 $self
195} 305}
196 306
197sub _shutdown { 307sub _shutdown {
198 my ($self) = @_; 308 my ($self) = @_;
199 309
310 delete $self->{_tw};
200 delete $self->{rw}; 311 delete $self->{_rw};
201 delete $self->{ww}; 312 delete $self->{_ww};
202 delete $self->{fh}; 313 delete $self->{fh};
203}
204 314
315 &_freetls;
316
317 delete $self->{on_read};
318 delete $self->{_queue};
319}
320
205sub error { 321sub _error {
206 my ($self) = @_; 322 my ($self, $errno, $fatal) = @_;
207 323
208 {
209 local $!;
210 $self->_shutdown; 324 $self->_shutdown
211 } 325 if $fatal;
326
327 $! = $errno;
212 328
213 if ($self->{on_error}) { 329 if ($self->{on_error}) {
214 $self->{on_error}($self); 330 $self->{on_error}($self, $fatal);
215 } else { 331 } elsif ($self->{fh}) {
216 die "AnyEvent::Handle uncaught fatal error: $!"; 332 Carp::croak "AnyEvent::Handle uncaught error: $!";
217 } 333 }
218} 334}
219 335
220=item $fh = $handle->fh 336=item $fh = $handle->fh
221 337
222This method returns the file handle of the L<AnyEvent::Handle> object. 338This method returns the file handle used to create the L<AnyEvent::Handle> object.
223 339
224=cut 340=cut
225 341
226sub fh { $_[0]->{fh} } 342sub fh { $_[0]{fh} }
227 343
228=item $handle->on_error ($cb) 344=item $handle->on_error ($cb)
229 345
230Replace the current C<on_error> callback (see the C<on_error> constructor argument). 346Replace the current C<on_error> callback (see the C<on_error> constructor argument).
231 347
243 359
244sub on_eof { 360sub on_eof {
245 $_[0]{on_eof} = $_[1]; 361 $_[0]{on_eof} = $_[1];
246} 362}
247 363
364=item $handle->on_timeout ($cb)
365
366Replace the current C<on_timeout> callback, or disables the callback (but
367not the timeout) if C<$cb> = C<undef>. See the C<timeout> constructor
368argument and method.
369
370=cut
371
372sub on_timeout {
373 $_[0]{on_timeout} = $_[1];
374}
375
376=item $handle->autocork ($boolean)
377
378Enables or disables the current autocork behaviour (see C<autocork>
379constructor argument).
380
381=cut
382
383=item $handle->no_delay ($boolean)
384
385Enables or disables the C<no_delay> setting (see constructor argument of
386the same name for details).
387
388=cut
389
390sub no_delay {
391 $_[0]{no_delay} = $_[1];
392
393 eval {
394 local $SIG{__DIE__};
395 setsockopt $_[0]{fh}, &Socket::IPPROTO_TCP, &Socket::TCP_NODELAY, int $_[1];
396 };
397}
398
399#############################################################################
400
401=item $handle->timeout ($seconds)
402
403Configures (or disables) the inactivity timeout.
404
405=cut
406
407sub timeout {
408 my ($self, $timeout) = @_;
409
410 $self->{timeout} = $timeout;
411 $self->_timeout;
412}
413
414# reset the timeout watcher, as neccessary
415# also check for time-outs
416sub _timeout {
417 my ($self) = @_;
418
419 if ($self->{timeout}) {
420 my $NOW = AnyEvent->now;
421
422 # when would the timeout trigger?
423 my $after = $self->{_activity} + $self->{timeout} - $NOW;
424
425 # now or in the past already?
426 if ($after <= 0) {
427 $self->{_activity} = $NOW;
428
429 if ($self->{on_timeout}) {
430 $self->{on_timeout}($self);
431 } else {
432 $self->_error (&Errno::ETIMEDOUT);
433 }
434
435 # callback could have changed timeout value, optimise
436 return unless $self->{timeout};
437
438 # calculate new after
439 $after = $self->{timeout};
440 }
441
442 Scalar::Util::weaken $self;
443 return unless $self; # ->error could have destroyed $self
444
445 $self->{_tw} ||= AnyEvent->timer (after => $after, cb => sub {
446 delete $self->{_tw};
447 $self->_timeout;
448 });
449 } else {
450 delete $self->{_tw};
451 }
452}
453
248############################################################################# 454#############################################################################
249 455
250=back 456=back
251 457
252=head2 WRITE QUEUE 458=head2 WRITE QUEUE
273 my ($self, $cb) = @_; 479 my ($self, $cb) = @_;
274 480
275 $self->{on_drain} = $cb; 481 $self->{on_drain} = $cb;
276 482
277 $cb->($self) 483 $cb->($self)
278 if $cb && $self->{low_water_mark} >= length $self->{wbuf}; 484 if $cb && $self->{low_water_mark} >= (length $self->{wbuf}) + (length $self->{_tls_wbuf});
279} 485}
280 486
281=item $handle->push_write ($data) 487=item $handle->push_write ($data)
282 488
283Queues the given scalar to be written. You can push as much data as you 489Queues the given scalar to be written. You can push as much data as you
287=cut 493=cut
288 494
289sub _drain_wbuf { 495sub _drain_wbuf {
290 my ($self) = @_; 496 my ($self) = @_;
291 497
292 unless ($self->{ww}) { 498 if (!$self->{_ww} && length $self->{wbuf}) {
499
293 Scalar::Util::weaken $self; 500 Scalar::Util::weaken $self;
501
294 my $cb = sub { 502 my $cb = sub {
295 my $len = syswrite $self->{fh}, $self->{wbuf}; 503 my $len = syswrite $self->{fh}, $self->{wbuf};
296 504
297 if ($len > 0) { 505 if ($len >= 0) {
298 substr $self->{wbuf}, 0, $len, ""; 506 substr $self->{wbuf}, 0, $len, "";
299 507
508 $self->{_activity} = AnyEvent->now;
509
300 $self->{on_drain}($self) 510 $self->{on_drain}($self)
301 if $self->{low_water_mark} >= length $self->{wbuf} 511 if $self->{low_water_mark} >= (length $self->{wbuf}) + (length $self->{_tls_wbuf})
302 && $self->{on_drain}; 512 && $self->{on_drain};
303 513
304 delete $self->{ww} unless length $self->{wbuf}; 514 delete $self->{_ww} unless length $self->{wbuf};
305 } elsif ($! != EAGAIN && $! != EINTR) { 515 } elsif ($! != EAGAIN && $! != EINTR && $! != WSAEWOULDBLOCK) {
306 $self->error; 516 $self->_error ($!, 1);
307 } 517 }
308 }; 518 };
309 519
520 # try to write data immediately
521 $cb->() unless $self->{autocork};
522
523 # if still data left in wbuf, we need to poll
310 $self->{ww} = AnyEvent->io (fh => $self->{fh}, poll => "w", cb => $cb); 524 $self->{_ww} = AnyEvent->io (fh => $self->{fh}, poll => "w", cb => $cb)
311 525 if length $self->{wbuf};
312 $cb->($self);
313 }; 526 };
527}
528
529our %WH;
530
531sub register_write_type($$) {
532 $WH{$_[0]} = $_[1];
314} 533}
315 534
316sub push_write { 535sub push_write {
317 my $self = shift; 536 my $self = shift;
318 537
538 if (@_ > 1) {
539 my $type = shift;
540
541 @_ = ($WH{$type} or Carp::croak "unsupported type passed to AnyEvent::Handle::push_write")
542 ->($self, @_);
543 }
544
319 if ($self->{filter_w}) { 545 if ($self->{tls}) {
320 $self->{filter_w}->($self, \$_[0]); 546 $self->{_tls_wbuf} .= $_[0];
547
548 &_dotls ($self);
321 } else { 549 } else {
322 $self->{wbuf} .= $_[0]; 550 $self->{wbuf} .= $_[0];
323 $self->_drain_wbuf; 551 $self->_drain_wbuf;
324 } 552 }
325} 553}
326 554
555=item $handle->push_write (type => @args)
556
557Instead of formatting your data yourself, you can also let this module do
558the job by specifying a type and type-specific arguments.
559
560Predefined types are (if you have ideas for additional types, feel free to
561drop by and tell us):
562
563=over 4
564
565=item netstring => $string
566
567Formats the given value as netstring
568(http://cr.yp.to/proto/netstrings.txt, this is not a recommendation to use them).
569
570=cut
571
572register_write_type netstring => sub {
573 my ($self, $string) = @_;
574
575 (length $string) . ":$string,"
576};
577
578=item packstring => $format, $data
579
580An octet string prefixed with an encoded length. The encoding C<$format>
581uses the same format as a Perl C<pack> format, but must specify a single
582integer only (only one of C<cCsSlLqQiInNvVjJw> is allowed, plus an
583optional C<!>, C<< < >> or C<< > >> modifier).
584
585=cut
586
587register_write_type packstring => sub {
588 my ($self, $format, $string) = @_;
589
590 pack "$format/a*", $string
591};
592
593=item json => $array_or_hashref
594
595Encodes the given hash or array reference into a JSON object. Unless you
596provide your own JSON object, this means it will be encoded to JSON text
597in UTF-8.
598
599JSON objects (and arrays) are self-delimiting, so you can write JSON at
600one end of a handle and read them at the other end without using any
601additional framing.
602
603The generated JSON text is guaranteed not to contain any newlines: While
604this module doesn't need delimiters after or between JSON texts to be
605able to read them, many other languages depend on that.
606
607A simple RPC protocol that interoperates easily with others is to send
608JSON arrays (or objects, although arrays are usually the better choice as
609they mimic how function argument passing works) and a newline after each
610JSON text:
611
612 $handle->push_write (json => ["method", "arg1", "arg2"]); # whatever
613 $handle->push_write ("\012");
614
615An AnyEvent::Handle receiver would simply use the C<json> read type and
616rely on the fact that the newline will be skipped as leading whitespace:
617
618 $handle->push_read (json => sub { my $array = $_[1]; ... });
619
620Other languages could read single lines terminated by a newline and pass
621this line into their JSON decoder of choice.
622
623=cut
624
625register_write_type json => sub {
626 my ($self, $ref) = @_;
627
628 require JSON;
629
630 $self->{json} ? $self->{json}->encode ($ref)
631 : JSON::encode_json ($ref)
632};
633
634=item storable => $reference
635
636Freezes the given reference using L<Storable> and writes it to the
637handle. Uses the C<nfreeze> format.
638
639=cut
640
641register_write_type storable => sub {
642 my ($self, $ref) = @_;
643
644 require Storable;
645
646 pack "w/a*", Storable::nfreeze ($ref)
647};
648
649=back
650
651=item AnyEvent::Handle::register_write_type type => $coderef->($handle, @args)
652
653This function (not method) lets you add your own types to C<push_write>.
654Whenever the given C<type> is used, C<push_write> will invoke the code
655reference with the handle object and the remaining arguments.
656
657The code reference is supposed to return a single octet string that will
658be appended to the write buffer.
659
660Note that this is a function, and all types registered this way will be
661global, so try to use unique names.
662
663=cut
664
327############################################################################# 665#############################################################################
328 666
329=back 667=back
330 668
331=head2 READ QUEUE 669=head2 READ QUEUE
337ways, the "simple" way, using only C<on_read> and the "complex" way, using 675ways, the "simple" way, using only C<on_read> and the "complex" way, using
338a queue. 676a queue.
339 677
340In the simple case, you just install an C<on_read> callback and whenever 678In the simple case, you just install an C<on_read> callback and whenever
341new data arrives, it will be called. You can then remove some data (if 679new data arrives, it will be called. You can then remove some data (if
342enough is there) from the read buffer (C<< $handle->rbuf >>) if you want 680enough is there) from the read buffer (C<< $handle->rbuf >>). Or you cna
343or not. 681leave the data there if you want to accumulate more (e.g. when only a
682partial message has been received so far).
344 683
345In the more complex case, you want to queue multiple callbacks. In this 684In the more complex case, you want to queue multiple callbacks. In this
346case, AnyEvent::Handle will call the first queued callback each time new 685case, AnyEvent::Handle will call the first queued callback each time new
347data arrives and removes it when it has done its job (see C<push_read>, 686data arrives (also the first time it is queued) and removes it when it has
348below). 687done its job (see C<push_read>, below).
349 688
350This way you can, for example, push three line-reads, followed by reading 689This way you can, for example, push three line-reads, followed by reading
351a chunk of data, and AnyEvent::Handle will execute them in order. 690a chunk of data, and AnyEvent::Handle will execute them in order.
352 691
353Example 1: EPP protocol parser. EPP sends 4 byte length info, followed by 692Example 1: EPP protocol parser. EPP sends 4 byte length info, followed by
354the specified number of bytes which give an XML datagram. 693the specified number of bytes which give an XML datagram.
355 694
356 # in the default state, expect some header bytes 695 # in the default state, expect some header bytes
357 $handle->on_read (sub { 696 $handle->on_read (sub {
358 # some data is here, now queue the length-header-read (4 octets) 697 # some data is here, now queue the length-header-read (4 octets)
359 shift->unshift_read_chunk (4, sub { 698 shift->unshift_read (chunk => 4, sub {
360 # header arrived, decode 699 # header arrived, decode
361 my $len = unpack "N", $_[1]; 700 my $len = unpack "N", $_[1];
362 701
363 # now read the payload 702 # now read the payload
364 shift->unshift_read_chunk ($len, sub { 703 shift->unshift_read (chunk => $len, sub {
365 my $xml = $_[1]; 704 my $xml = $_[1];
366 # handle xml 705 # handle xml
367 }); 706 });
368 }); 707 });
369 }); 708 });
370 709
371Example 2: Implement a client for a protocol that replies either with 710Example 2: Implement a client for a protocol that replies either with "OK"
372"OK" and another line or "ERROR" for one request, and 64 bytes for the 711and another line or "ERROR" for the first request that is sent, and 64
373second request. Due tot he availability of a full queue, we can just 712bytes for the second request. Due to the availability of a queue, we can
374pipeline sending both requests and manipulate the queue as necessary in 713just pipeline sending both requests and manipulate the queue as necessary
375the callbacks: 714in the callbacks.
376 715
377 # request one 716When the first callback is called and sees an "OK" response, it will
717C<unshift> another line-read. This line-read will be queued I<before> the
71864-byte chunk callback.
719
720 # request one, returns either "OK + extra line" or "ERROR"
378 $handle->push_write ("request 1\015\012"); 721 $handle->push_write ("request 1\015\012");
379 722
380 # we expect "ERROR" or "OK" as response, so push a line read 723 # we expect "ERROR" or "OK" as response, so push a line read
381 $handle->push_read_line (sub { 724 $handle->push_read (line => sub {
382 # if we got an "OK", we have to _prepend_ another line, 725 # if we got an "OK", we have to _prepend_ another line,
383 # so it will be read before the second request reads its 64 bytes 726 # so it will be read before the second request reads its 64 bytes
384 # which are already in the queue when this callback is called 727 # which are already in the queue when this callback is called
385 # we don't do this in case we got an error 728 # we don't do this in case we got an error
386 if ($_[1] eq "OK") { 729 if ($_[1] eq "OK") {
387 $_[0]->unshift_read_line (sub { 730 $_[0]->unshift_read (line => sub {
388 my $response = $_[1]; 731 my $response = $_[1];
389 ... 732 ...
390 }); 733 });
391 } 734 }
392 }); 735 });
393 736
394 # request two 737 # request two, simply returns 64 octets
395 $handle->push_write ("request 2\015\012"); 738 $handle->push_write ("request 2\015\012");
396 739
397 # simply read 64 bytes, always 740 # simply read 64 bytes, always
398 $handle->push_read_chunk (64, sub { 741 $handle->push_read (chunk => 64, sub {
399 my $response = $_[1]; 742 my $response = $_[1];
400 ... 743 ...
401 }); 744 });
402 745
403=over 4 746=over 4
404 747
405=cut 748=cut
406 749
407sub _drain_rbuf { 750sub _drain_rbuf {
408 my ($self) = @_; 751 my ($self) = @_;
752
753 local $self->{_in_drain} = 1;
409 754
410 if ( 755 if (
411 defined $self->{rbuf_max} 756 defined $self->{rbuf_max}
412 && $self->{rbuf_max} < length $self->{rbuf} 757 && $self->{rbuf_max} < length $self->{rbuf}
413 ) { 758 ) {
414 $! = &Errno::ENOSPC; return $self->error; 759 $self->_error (&Errno::ENOSPC, 1), return;
415 } 760 }
416 761
417 return if $self->{in_drain}; 762 while () {
418 local $self->{in_drain} = 1;
419
420 while (my $len = length $self->{rbuf}) { 763 my $len = length $self->{rbuf};
421 no strict 'refs'; 764
422 if (my $cb = shift @{ $self->{queue} }) { 765 if (my $cb = shift @{ $self->{_queue} }) {
423 if (!$cb->($self)) { 766 unless ($cb->($self)) {
424 if ($self->{eof}) { 767 if ($self->{_eof}) {
425 # no progress can be made (not enough data and no data forthcoming) 768 # no progress can be made (not enough data and no data forthcoming)
426 $! = &Errno::EPIPE; return $self->error; 769 $self->_error (&Errno::EPIPE, 1), return;
427 } 770 }
428 771
429 unshift @{ $self->{queue} }, $cb; 772 unshift @{ $self->{_queue} }, $cb;
430 return; 773 last;
431 } 774 }
432 } elsif ($self->{on_read}) { 775 } elsif ($self->{on_read}) {
776 last unless $len;
777
433 $self->{on_read}($self); 778 $self->{on_read}($self);
434 779
435 if ( 780 if (
436 $self->{eof} # if no further data will arrive
437 && $len == length $self->{rbuf} # and no data has been consumed 781 $len == length $self->{rbuf} # if no data has been consumed
438 && !@{ $self->{queue} } # and the queue is still empty 782 && !@{ $self->{_queue} } # and the queue is still empty
439 && $self->{on_read} # and we still want to read data 783 && $self->{on_read} # but we still have on_read
440 ) { 784 ) {
785 # no further data will arrive
441 # then no progress can be made 786 # so no progress can be made
442 $! = &Errno::EPIPE; return $self->error; 787 $self->_error (&Errno::EPIPE, 1), return
788 if $self->{_eof};
789
790 last; # more data might arrive
443 } 791 }
444 } else { 792 } else {
445 # read side becomes idle 793 # read side becomes idle
446 delete $self->{rw}; 794 delete $self->{_rw} unless $self->{tls};
447 return; 795 last;
448 } 796 }
449 } 797 }
450 798
451 if ($self->{eof}) { 799 if ($self->{_eof}) {
452 $self->_shutdown; 800 if ($self->{on_eof}) {
453 $self->{on_eof}($self) 801 $self->{on_eof}($self)
454 if $self->{on_eof}; 802 } else {
803 $self->_error (0, 1);
804 }
805 }
806
807 # may need to restart read watcher
808 unless ($self->{_rw}) {
809 $self->start_read
810 if $self->{on_read} || @{ $self->{_queue} };
455 } 811 }
456} 812}
457 813
458=item $handle->on_read ($cb) 814=item $handle->on_read ($cb)
459 815
465 821
466sub on_read { 822sub on_read {
467 my ($self, $cb) = @_; 823 my ($self, $cb) = @_;
468 824
469 $self->{on_read} = $cb; 825 $self->{on_read} = $cb;
826 $self->_drain_rbuf if $cb && !$self->{_in_drain};
470} 827}
471 828
472=item $handle->rbuf 829=item $handle->rbuf
473 830
474Returns the read buffer (as a modifiable lvalue). 831Returns the read buffer (as a modifiable lvalue).
505interested in (which can be none at all) and return a true value. After returning 862interested in (which can be none at all) and return a true value. After returning
506true, it will be removed from the queue. 863true, it will be removed from the queue.
507 864
508=cut 865=cut
509 866
867our %RH;
868
869sub register_read_type($$) {
870 $RH{$_[0]} = $_[1];
871}
872
510sub push_read { 873sub push_read {
511 my ($self, $cb) = @_; 874 my $self = shift;
875 my $cb = pop;
512 876
877 if (@_) {
878 my $type = shift;
879
880 $cb = ($RH{$type} or Carp::croak "unsupported type passed to AnyEvent::Handle::push_read")
881 ->($self, $cb, @_);
882 }
883
513 push @{ $self->{queue} }, $cb; 884 push @{ $self->{_queue} }, $cb;
514 $self->_drain_rbuf; 885 $self->_drain_rbuf unless $self->{_in_drain};
515} 886}
516 887
517sub unshift_read { 888sub unshift_read {
518 my ($self, $cb) = @_; 889 my $self = shift;
890 my $cb = pop;
519 891
892 if (@_) {
893 my $type = shift;
894
895 $cb = ($RH{$type} or Carp::croak "unsupported type passed to AnyEvent::Handle::unshift_read")
896 ->($self, $cb, @_);
897 }
898
899
520 push @{ $self->{queue} }, $cb; 900 unshift @{ $self->{_queue} }, $cb;
521 $self->_drain_rbuf; 901 $self->_drain_rbuf unless $self->{_in_drain};
522} 902}
523 903
524=item $handle->push_read_chunk ($len, $cb->($self, $data)) 904=item $handle->push_read (type => @args, $cb)
525 905
526=item $handle->unshift_read_chunk ($len, $cb->($self, $data)) 906=item $handle->unshift_read (type => @args, $cb)
527 907
528Append the given callback to the end of the queue (C<push_read_chunk>) or 908Instead of providing a callback that parses the data itself you can chose
529prepend it (C<unshift_read_chunk>). 909between a number of predefined parsing formats, for chunks of data, lines
910etc.
530 911
531The callback will be called only once C<$len> bytes have been read, and 912Predefined types are (if you have ideas for additional types, feel free to
532these C<$len> bytes will be passed to the callback. 913drop by and tell us):
533 914
534=cut 915=over 4
535 916
536sub _read_chunk($$) { 917=item chunk => $octets, $cb->($handle, $data)
918
919Invoke the callback only once C<$octets> bytes have been read. Pass the
920data read to the callback. The callback will never be called with less
921data.
922
923Example: read 2 bytes.
924
925 $handle->push_read (chunk => 2, sub {
926 warn "yay ", unpack "H*", $_[1];
927 });
928
929=cut
930
931register_read_type chunk => sub {
537 my ($self, $len, $cb) = @_; 932 my ($self, $cb, $len) = @_;
538 933
539 sub { 934 sub {
540 $len <= length $_[0]{rbuf} or return; 935 $len <= length $_[0]{rbuf} or return;
541 $cb->($_[0], substr $_[0]{rbuf}, 0, $len, ""); 936 $cb->($_[0], substr $_[0]{rbuf}, 0, $len, "");
542 1 937 1
543 } 938 }
544} 939};
545 940
546sub push_read_chunk { 941=item line => [$eol, ]$cb->($handle, $line, $eol)
547 $_[0]->push_read (&_read_chunk);
548}
549
550
551sub unshift_read_chunk {
552 $_[0]->unshift_read (&_read_chunk);
553}
554
555=item $handle->push_read_line ([$eol, ]$cb->($self, $line, $eol))
556
557=item $handle->unshift_read_line ([$eol, ]$cb->($self, $line, $eol))
558
559Append the given callback to the end of the queue (C<push_read_line>) or
560prepend it (C<unshift_read_line>).
561 942
562The callback will be called only once a full line (including the end of 943The callback will be called only once a full line (including the end of
563line marker, C<$eol>) has been read. This line (excluding the end of line 944line marker, C<$eol>) has been read. This line (excluding the end of line
564marker) will be passed to the callback as second argument (C<$line>), and 945marker) will be passed to the callback as second argument (C<$line>), and
565the end of line marker as the third argument (C<$eol>). 946the end of line marker as the third argument (C<$eol>).
576Partial lines at the end of the stream will never be returned, as they are 957Partial lines at the end of the stream will never be returned, as they are
577not marked by the end of line marker. 958not marked by the end of line marker.
578 959
579=cut 960=cut
580 961
581sub _read_line($$) { 962register_read_type line => sub {
582 my $self = shift; 963 my ($self, $cb, $eol) = @_;
583 my $cb = pop;
584 my $eol = @_ ? shift : qr|(\015?\012)|;
585 my $pos;
586 964
965 if (@_ < 3) {
966 # this is more than twice as fast as the generic code below
967 sub {
968 $_[0]{rbuf} =~ s/^([^\015\012]*)(\015?\012)// or return;
969
970 $cb->($_[0], $1, $2);
971 1
972 }
973 } else {
587 $eol = quotemeta $eol unless ref $eol; 974 $eol = quotemeta $eol unless ref $eol;
588 $eol = qr|^(.*?)($eol)|s; 975 $eol = qr|^(.*?)($eol)|s;
976
977 sub {
978 $_[0]{rbuf} =~ s/$eol// or return;
979
980 $cb->($_[0], $1, $2);
981 1
982 }
983 }
984};
985
986=item regex => $accept[, $reject[, $skip], $cb->($handle, $data)
987
988Makes a regex match against the regex object C<$accept> and returns
989everything up to and including the match.
990
991Example: read a single line terminated by '\n'.
992
993 $handle->push_read (regex => qr<\n>, sub { ... });
994
995If C<$reject> is given and not undef, then it determines when the data is
996to be rejected: it is matched against the data when the C<$accept> regex
997does not match and generates an C<EBADMSG> error when it matches. This is
998useful to quickly reject wrong data (to avoid waiting for a timeout or a
999receive buffer overflow).
1000
1001Example: expect a single decimal number followed by whitespace, reject
1002anything else (not the use of an anchor).
1003
1004 $handle->push_read (regex => qr<^[0-9]+\s>, qr<[^0-9]>, sub { ... });
1005
1006If C<$skip> is given and not C<undef>, then it will be matched against
1007the receive buffer when neither C<$accept> nor C<$reject> match,
1008and everything preceding and including the match will be accepted
1009unconditionally. This is useful to skip large amounts of data that you
1010know cannot be matched, so that the C<$accept> or C<$reject> regex do not
1011have to start matching from the beginning. This is purely an optimisation
1012and is usually worth only when you expect more than a few kilobytes.
1013
1014Example: expect a http header, which ends at C<\015\012\015\012>. Since we
1015expect the header to be very large (it isn't in practise, but...), we use
1016a skip regex to skip initial portions. The skip regex is tricky in that
1017it only accepts something not ending in either \015 or \012, as these are
1018required for the accept regex.
1019
1020 $handle->push_read (regex =>
1021 qr<\015\012\015\012>,
1022 undef, # no reject
1023 qr<^.*[^\015\012]>,
1024 sub { ... });
1025
1026=cut
1027
1028register_read_type regex => sub {
1029 my ($self, $cb, $accept, $reject, $skip) = @_;
1030
1031 my $data;
1032 my $rbuf = \$self->{rbuf};
589 1033
590 sub { 1034 sub {
591 $_[0]{rbuf} =~ s/$eol// or return; 1035 # accept
1036 if ($$rbuf =~ $accept) {
1037 $data .= substr $$rbuf, 0, $+[0], "";
1038 $cb->($self, $data);
1039 return 1;
1040 }
1041
1042 # reject
1043 if ($reject && $$rbuf =~ $reject) {
1044 $self->_error (&Errno::EBADMSG);
1045 }
592 1046
593 $cb->($_[0], $1, $2); 1047 # skip
1048 if ($skip && $$rbuf =~ $skip) {
1049 $data .= substr $$rbuf, 0, $+[0], "";
1050 }
1051
1052 ()
1053 }
1054};
1055
1056=item netstring => $cb->($handle, $string)
1057
1058A netstring (http://cr.yp.to/proto/netstrings.txt, this is not an endorsement).
1059
1060Throws an error with C<$!> set to EBADMSG on format violations.
1061
1062=cut
1063
1064register_read_type netstring => sub {
1065 my ($self, $cb) = @_;
1066
1067 sub {
1068 unless ($_[0]{rbuf} =~ s/^(0|[1-9][0-9]*)://) {
1069 if ($_[0]{rbuf} =~ /[^0-9]/) {
1070 $self->_error (&Errno::EBADMSG);
1071 }
1072 return;
1073 }
1074
1075 my $len = $1;
1076
1077 $self->unshift_read (chunk => $len, sub {
1078 my $string = $_[1];
1079 $_[0]->unshift_read (chunk => 1, sub {
1080 if ($_[1] eq ",") {
1081 $cb->($_[0], $string);
1082 } else {
1083 $self->_error (&Errno::EBADMSG);
1084 }
1085 });
1086 });
1087
594 1 1088 1
595 } 1089 }
596} 1090};
597 1091
598sub push_read_line { 1092=item packstring => $format, $cb->($handle, $string)
599 $_[0]->push_read (&_read_line);
600}
601 1093
602sub unshift_read_line { 1094An octet string prefixed with an encoded length. The encoding C<$format>
603 $_[0]->unshift_read (&_read_line); 1095uses the same format as a Perl C<pack> format, but must specify a single
604} 1096integer only (only one of C<cCsSlLqQiInNvVjJw> is allowed, plus an
1097optional C<!>, C<< < >> or C<< > >> modifier).
1098
1099For example, DNS over TCP uses a prefix of C<n> (2 octet network order),
1100EPP uses a prefix of C<N> (4 octtes).
1101
1102Example: read a block of data prefixed by its length in BER-encoded
1103format (very efficient).
1104
1105 $handle->push_read (packstring => "w", sub {
1106 my ($handle, $data) = @_;
1107 });
1108
1109=cut
1110
1111register_read_type packstring => sub {
1112 my ($self, $cb, $format) = @_;
1113
1114 sub {
1115 # when we can use 5.10 we can use ".", but for 5.8 we use the re-pack method
1116 defined (my $len = eval { unpack $format, $_[0]{rbuf} })
1117 or return;
1118
1119 $format = length pack $format, $len;
1120
1121 # bypass unshift if we already have the remaining chunk
1122 if ($format + $len <= length $_[0]{rbuf}) {
1123 my $data = substr $_[0]{rbuf}, $format, $len;
1124 substr $_[0]{rbuf}, 0, $format + $len, "";
1125 $cb->($_[0], $data);
1126 } else {
1127 # remove prefix
1128 substr $_[0]{rbuf}, 0, $format, "";
1129
1130 # read remaining chunk
1131 $_[0]->unshift_read (chunk => $len, $cb);
1132 }
1133
1134 1
1135 }
1136};
1137
1138=item json => $cb->($handle, $hash_or_arrayref)
1139
1140Reads a JSON object or array, decodes it and passes it to the callback.
1141
1142If a C<json> object was passed to the constructor, then that will be used
1143for the final decode, otherwise it will create a JSON coder expecting UTF-8.
1144
1145This read type uses the incremental parser available with JSON version
11462.09 (and JSON::XS version 2.2) and above. You have to provide a
1147dependency on your own: this module will load the JSON module, but
1148AnyEvent does not depend on it itself.
1149
1150Since JSON texts are fully self-delimiting, the C<json> read and write
1151types are an ideal simple RPC protocol: just exchange JSON datagrams. See
1152the C<json> write type description, above, for an actual example.
1153
1154=cut
1155
1156register_read_type json => sub {
1157 my ($self, $cb) = @_;
1158
1159 require JSON;
1160
1161 my $data;
1162 my $rbuf = \$self->{rbuf};
1163
1164 my $json = $self->{json} ||= JSON->new->utf8;
1165
1166 sub {
1167 my $ref = $json->incr_parse ($self->{rbuf});
1168
1169 if ($ref) {
1170 $self->{rbuf} = $json->incr_text;
1171 $json->incr_text = "";
1172 $cb->($self, $ref);
1173
1174 1
1175 } else {
1176 $self->{rbuf} = "";
1177 ()
1178 }
1179 }
1180};
1181
1182=item storable => $cb->($handle, $ref)
1183
1184Deserialises a L<Storable> frozen representation as written by the
1185C<storable> write type (BER-encoded length prefix followed by nfreeze'd
1186data).
1187
1188Raises C<EBADMSG> error if the data could not be decoded.
1189
1190=cut
1191
1192register_read_type storable => sub {
1193 my ($self, $cb) = @_;
1194
1195 require Storable;
1196
1197 sub {
1198 # when we can use 5.10 we can use ".", but for 5.8 we use the re-pack method
1199 defined (my $len = eval { unpack "w", $_[0]{rbuf} })
1200 or return;
1201
1202 my $format = length pack "w", $len;
1203
1204 # bypass unshift if we already have the remaining chunk
1205 if ($format + $len <= length $_[0]{rbuf}) {
1206 my $data = substr $_[0]{rbuf}, $format, $len;
1207 substr $_[0]{rbuf}, 0, $format + $len, "";
1208 $cb->($_[0], Storable::thaw ($data));
1209 } else {
1210 # remove prefix
1211 substr $_[0]{rbuf}, 0, $format, "";
1212
1213 # read remaining chunk
1214 $_[0]->unshift_read (chunk => $len, sub {
1215 if (my $ref = eval { Storable::thaw ($_[1]) }) {
1216 $cb->($_[0], $ref);
1217 } else {
1218 $self->_error (&Errno::EBADMSG);
1219 }
1220 });
1221 }
1222
1223 1
1224 }
1225};
1226
1227=back
1228
1229=item AnyEvent::Handle::register_read_type type => $coderef->($handle, $cb, @args)
1230
1231This function (not method) lets you add your own types to C<push_read>.
1232
1233Whenever the given C<type> is used, C<push_read> will invoke the code
1234reference with the handle object, the callback and the remaining
1235arguments.
1236
1237The code reference is supposed to return a callback (usually a closure)
1238that works as a plain read callback (see C<< ->push_read ($cb) >>).
1239
1240It should invoke the passed callback when it is done reading (remember to
1241pass C<$handle> as first argument as all other callbacks do that).
1242
1243Note that this is a function, and all types registered this way will be
1244global, so try to use unique names.
1245
1246For examples, see the source of this module (F<perldoc -m AnyEvent::Handle>,
1247search for C<register_read_type>)).
605 1248
606=item $handle->stop_read 1249=item $handle->stop_read
607 1250
608=item $handle->start_read 1251=item $handle->start_read
609 1252
610In rare cases you actually do not want to read anything from the 1253In rare cases you actually do not want to read anything from the
611socket. In this case you can call C<stop_read>. Neither C<on_read> no 1254socket. In this case you can call C<stop_read>. Neither C<on_read> nor
612any queued callbacks will be executed then. To start reading again, call 1255any queued callbacks will be executed then. To start reading again, call
613C<start_read>. 1256C<start_read>.
614 1257
1258Note that AnyEvent::Handle will automatically C<start_read> for you when
1259you change the C<on_read> callback or push/unshift a read callback, and it
1260will automatically C<stop_read> for you when neither C<on_read> is set nor
1261there are any read requests in the queue.
1262
1263These methods will have no effect when in TLS mode (as TLS doesn't support
1264half-duplex connections).
1265
615=cut 1266=cut
616 1267
617sub stop_read { 1268sub stop_read {
618 my ($self) = @_; 1269 my ($self) = @_;
619 1270
620 delete $self->{rw}; 1271 delete $self->{_rw} unless $self->{tls};
621} 1272}
622 1273
623sub start_read { 1274sub start_read {
624 my ($self) = @_; 1275 my ($self) = @_;
625 1276
626 unless ($self->{rw} || $self->{eof}) { 1277 unless ($self->{_rw} || $self->{_eof}) {
627 Scalar::Util::weaken $self; 1278 Scalar::Util::weaken $self;
628 1279
629 $self->{rw} = AnyEvent->io (fh => $self->{fh}, poll => "r", cb => sub { 1280 $self->{_rw} = AnyEvent->io (fh => $self->{fh}, poll => "r", cb => sub {
630 my $rbuf = $self->{filter_r} ? \my $buf : \$self->{rbuf}; 1281 my $rbuf = \($self->{tls} ? my $buf : $self->{rbuf});
631 my $len = sysread $self->{fh}, $$rbuf, $self->{read_size} || 8192, length $$rbuf; 1282 my $len = sysread $self->{fh}, $$rbuf, $self->{read_size} || 8192, length $$rbuf;
632 1283
633 if ($len > 0) { 1284 if ($len > 0) {
634 $self->{filter_r} 1285 $self->{_activity} = AnyEvent->now;
635 ? $self->{filter_r}->($self, $rbuf) 1286
636 : $self->_drain_rbuf; 1287 if ($self->{tls}) {
1288 Net::SSLeay::BIO_write ($self->{_rbio}, $$rbuf);
1289
1290 &_dotls ($self);
1291 } else {
1292 $self->_drain_rbuf unless $self->{_in_drain};
1293 }
637 1294
638 } elsif (defined $len) { 1295 } elsif (defined $len) {
639 delete $self->{rw}; 1296 delete $self->{_rw};
640 $self->{eof} = 1; 1297 $self->{_eof} = 1;
641 $self->_drain_rbuf; 1298 $self->_drain_rbuf unless $self->{_in_drain};
642 1299
643 } elsif ($! != EAGAIN && $! != EINTR) { 1300 } elsif ($! != EAGAIN && $! != EINTR && $! != WSAEWOULDBLOCK) {
644 return $self->error; 1301 return $self->_error ($!, 1);
645 } 1302 }
646 }); 1303 });
647 } 1304 }
648} 1305}
649 1306
1307# poll the write BIO and send the data if applicable
650sub _dotls { 1308sub _dotls {
651 my ($self) = @_; 1309 my ($self) = @_;
652 1310
1311 my $tmp;
1312
653 if (length $self->{tls_wbuf}) { 1313 if (length $self->{_tls_wbuf}) {
654 while ((my $len = Net::SSLeay::write ($self->{tls}, $self->{tls_wbuf})) > 0) { 1314 while (($tmp = Net::SSLeay::write ($self->{tls}, $self->{_tls_wbuf})) > 0) {
655 substr $self->{tls_wbuf}, 0, $len, ""; 1315 substr $self->{_tls_wbuf}, 0, $tmp, "";
656 } 1316 }
657 } 1317 }
658 1318
659 if (defined (my $buf = Net::SSLeay::BIO_read ($self->{tls_wbio}))) { 1319 while (defined ($tmp = Net::SSLeay::read ($self->{tls}))) {
1320 unless (length $tmp) {
1321 # let's treat SSL-eof as we treat normal EOF
1322 delete $self->{_rw};
1323 $self->{_eof} = 1;
1324 &_freetls;
1325 }
1326
1327 $self->{rbuf} .= $tmp;
1328 $self->_drain_rbuf unless $self->{_in_drain};
1329 $self->{tls} or return; # tls session might have gone away in callback
1330 }
1331
1332 $tmp = Net::SSLeay::get_error ($self->{tls}, -1);
1333
1334 if ($tmp != Net::SSLeay::ERROR_WANT_READ ()) {
1335 if ($tmp == Net::SSLeay::ERROR_SYSCALL ()) {
1336 return $self->_error ($!, 1);
1337 } elsif ($tmp == Net::SSLeay::ERROR_SSL ()) {
1338 return $self->_error (&Errno::EIO, 1);
1339 }
1340
1341 # all other errors are fine for our purposes
1342 }
1343
1344 while (length ($tmp = Net::SSLeay::BIO_read ($self->{_wbio}))) {
660 $self->{wbuf} .= $buf; 1345 $self->{wbuf} .= $tmp;
661 $self->_drain_wbuf; 1346 $self->_drain_wbuf;
662 }
663
664 while (defined (my $buf = Net::SSLeay::read ($self->{tls}))) {
665 $self->{rbuf} .= $buf;
666 $self->_drain_rbuf;
667 }
668
669 my $err = Net::SSLeay::get_error ($self->{tls}, -1);
670
671 if ($err!= Net::SSLeay::ERROR_WANT_READ ()) {
672 if ($err == Net::SSLeay::ERROR_SYSCALL ()) {
673 $self->error;
674 } elsif ($err == Net::SSLeay::ERROR_SSL ()) {
675 $! = &Errno::EIO;
676 $self->error;
677 }
678
679 # all others are fine for our purposes
680 } 1347 }
681} 1348}
682 1349
683=item $handle->starttls ($tls[, $tls_ctx]) 1350=item $handle->starttls ($tls[, $tls_ctx])
684 1351
690C<"connect">, C<"accept"> or an existing Net::SSLeay object). 1357C<"connect">, C<"accept"> or an existing Net::SSLeay object).
691 1358
692The second argument is the optional C<Net::SSLeay::CTX> object that is 1359The second argument is the optional C<Net::SSLeay::CTX> object that is
693used when AnyEvent::Handle has to create its own TLS connection object. 1360used when AnyEvent::Handle has to create its own TLS connection object.
694 1361
695=cut 1362The TLS connection object will end up in C<< $handle->{tls} >> after this
1363call and can be used or changed to your liking. Note that the handshake
1364might have already started when this function returns.
696 1365
697# TODO: maybe document... 1366If it an error to start a TLS handshake more than once per
1367AnyEvent::Handle object (this is due to bugs in OpenSSL).
1368
1369=cut
1370
698sub starttls { 1371sub starttls {
699 my ($self, $ssl, $ctx) = @_; 1372 my ($self, $ssl, $ctx) = @_;
700 1373
701 $self->stoptls; 1374 require Net::SSLeay;
702 1375
1376 Carp::croak "it is an error to call starttls more than once on an AnyEvent::Handle object"
1377 if $self->{tls};
1378
703 if ($ssl eq "accept") { 1379 if ($ssl eq "accept") {
704 $ssl = Net::SSLeay::new ($ctx || TLS_CTX ()); 1380 $ssl = Net::SSLeay::new ($ctx || TLS_CTX ());
705 Net::SSLeay::set_accept_state ($ssl); 1381 Net::SSLeay::set_accept_state ($ssl);
706 } elsif ($ssl eq "connect") { 1382 } elsif ($ssl eq "connect") {
707 $ssl = Net::SSLeay::new ($ctx || TLS_CTX ()); 1383 $ssl = Net::SSLeay::new ($ctx || TLS_CTX ());
713 # basically, this is deep magic (because SSL_read should have the same issues) 1389 # basically, this is deep magic (because SSL_read should have the same issues)
714 # but the openssl maintainers basically said: "trust us, it just works". 1390 # but the openssl maintainers basically said: "trust us, it just works".
715 # (unfortunately, we have to hardcode constants because the abysmally misdesigned 1391 # (unfortunately, we have to hardcode constants because the abysmally misdesigned
716 # and mismaintained ssleay-module doesn't even offer them). 1392 # and mismaintained ssleay-module doesn't even offer them).
717 # http://www.mail-archive.com/openssl-dev@openssl.org/msg22420.html 1393 # http://www.mail-archive.com/openssl-dev@openssl.org/msg22420.html
1394 #
1395 # in short: this is a mess.
1396 #
1397 # note that we do not try to keep the length constant between writes as we are required to do.
1398 # we assume that most (but not all) of this insanity only applies to non-blocking cases,
1399 # and we drive openssl fully in blocking mode here. Or maybe we don't - openssl seems to
1400 # have identity issues in that area.
718 Net::SSLeay::CTX_set_mode ($self->{tls}, 1401 Net::SSLeay::CTX_set_mode ($self->{tls},
719 (eval { Net::SSLeay::MODE_ENABLE_PARTIAL_WRITE () } || 1) 1402 (eval { local $SIG{__DIE__}; Net::SSLeay::MODE_ENABLE_PARTIAL_WRITE () } || 1)
720 | (eval { Net::SSLeay::MODE_ACCEPT_MOVING_WRITE_BUFFER () } || 2)); 1403 | (eval { local $SIG{__DIE__}; Net::SSLeay::MODE_ACCEPT_MOVING_WRITE_BUFFER () } || 2));
721 1404
722 $self->{tls_rbio} = Net::SSLeay::BIO_new (Net::SSLeay::BIO_s_mem ()); 1405 $self->{_rbio} = Net::SSLeay::BIO_new (Net::SSLeay::BIO_s_mem ());
723 $self->{tls_wbio} = Net::SSLeay::BIO_new (Net::SSLeay::BIO_s_mem ()); 1406 $self->{_wbio} = Net::SSLeay::BIO_new (Net::SSLeay::BIO_s_mem ());
724 1407
725 Net::SSLeay::set_bio ($ssl, $self->{tls_rbio}, $self->{tls_wbio}); 1408 Net::SSLeay::set_bio ($ssl, $self->{_rbio}, $self->{_wbio});
726 1409
727 $self->{filter_w} = sub { 1410 &_dotls; # need to trigger the initial handshake
728 $_[0]{tls_wbuf} .= ${$_[1]}; 1411 $self->start_read; # make sure we actually do read
729 &_dotls;
730 };
731 $self->{filter_r} = sub {
732 Net::SSLeay::BIO_write ($_[0]{tls_rbio}, ${$_[1]});
733 &_dotls;
734 };
735} 1412}
736 1413
737=item $handle->stoptls 1414=item $handle->stoptls
738 1415
739Destroys the SSL connection, if any. Partial read or write data will be 1416Shuts down the SSL connection - this makes a proper EOF handshake by
740lost. 1417sending a close notify to the other side, but since OpenSSL doesn't
1418support non-blocking shut downs, it is not possible to re-use the stream
1419afterwards.
741 1420
742=cut 1421=cut
743 1422
744sub stoptls { 1423sub stoptls {
745 my ($self) = @_; 1424 my ($self) = @_;
746 1425
1426 if ($self->{tls}) {
1427 Net::SSLeay::shutdown ($self->{tls});
1428
1429 &_dotls;
1430
1431 # we don't give a shit. no, we do, but we can't. no...
1432 # we, we... have to use openssl :/
1433 &_freetls;
1434 }
1435}
1436
1437sub _freetls {
1438 my ($self) = @_;
1439
1440 return unless $self->{tls};
1441
747 Net::SSLeay::free (delete $self->{tls}) if $self->{tls}; 1442 Net::SSLeay::free (delete $self->{tls});
748 delete $self->{tls_rbio}; 1443
749 delete $self->{tls_wbio}; 1444 delete @$self{qw(_rbio _wbio _tls_wbuf)};
750 delete $self->{tls_wbuf};
751 delete $self->{filter_r};
752 delete $self->{filter_w};
753} 1445}
754 1446
755sub DESTROY { 1447sub DESTROY {
756 my $self = shift; 1448 my $self = shift;
757 1449
758 $self->stoptls; 1450 &_freetls;
1451
1452 my $linger = exists $self->{linger} ? $self->{linger} : 3600;
1453
1454 if ($linger && length $self->{wbuf}) {
1455 my $fh = delete $self->{fh};
1456 my $wbuf = delete $self->{wbuf};
1457
1458 my @linger;
1459
1460 push @linger, AnyEvent->io (fh => $fh, poll => "w", cb => sub {
1461 my $len = syswrite $fh, $wbuf, length $wbuf;
1462
1463 if ($len > 0) {
1464 substr $wbuf, 0, $len, "";
1465 } else {
1466 @linger = (); # end
1467 }
1468 });
1469 push @linger, AnyEvent->timer (after => $linger, cb => sub {
1470 @linger = ();
1471 });
1472 }
1473}
1474
1475=item $handle->destroy
1476
1477Shuts down the handle object as much as possible - this call ensures that
1478no further callbacks will be invoked and resources will be freed as much
1479as possible. You must not call any methods on the object afterwards.
1480
1481Normally, you can just "forget" any references to an AnyEvent::Handle
1482object and it will simply shut down. This works in fatal error and EOF
1483callbacks, as well as code outside. It does I<NOT> work in a read or write
1484callback, so when you want to destroy the AnyEvent::Handle object from
1485within such an callback. You I<MUST> call C<< ->destroy >> explicitly in
1486that case.
1487
1488The handle might still linger in the background and write out remaining
1489data, as specified by the C<linger> option, however.
1490
1491=cut
1492
1493sub destroy {
1494 my ($self) = @_;
1495
1496 $self->DESTROY;
1497 %$self = ();
759} 1498}
760 1499
761=item AnyEvent::Handle::TLS_CTX 1500=item AnyEvent::Handle::TLS_CTX
762 1501
763This function creates and returns the Net::SSLeay::CTX object used by 1502This function creates and returns the Net::SSLeay::CTX object used by
793 } 1532 }
794} 1533}
795 1534
796=back 1535=back
797 1536
1537
1538=head1 NONFREQUENTLY ASKED QUESTIONS
1539
1540=over 4
1541
1542=item I C<undef> the AnyEvent::Handle reference inside my callback and
1543still get further invocations!
1544
1545That's because AnyEvent::Handle keeps a reference to itself when handling
1546read or write callbacks.
1547
1548It is only safe to "forget" the reference inside EOF or error callbacks,
1549from within all other callbacks, you need to explicitly call the C<<
1550->destroy >> method.
1551
1552=item I get different callback invocations in TLS mode/Why can't I pause
1553reading?
1554
1555Unlike, say, TCP, TLS connections do not consist of two independent
1556communication channels, one for each direction. Or put differently. The
1557read and write directions are not independent of each other: you cannot
1558write data unless you are also prepared to read, and vice versa.
1559
1560This can mean than, in TLS mode, you might get C<on_error> or C<on_eof>
1561callback invocations when you are not expecting any read data - the reason
1562is that AnyEvent::Handle always reads in TLS mode.
1563
1564During the connection, you have to make sure that you always have a
1565non-empty read-queue, or an C<on_read> watcher. At the end of the
1566connection (or when you no longer want to use it) you can call the
1567C<destroy> method.
1568
1569=item How do I read data until the other side closes the connection?
1570
1571If you just want to read your data into a perl scalar, the easiest way
1572to achieve this is by setting an C<on_read> callback that does nothing,
1573clearing the C<on_eof> callback and in the C<on_error> callback, the data
1574will be in C<$_[0]{rbuf}>:
1575
1576 $handle->on_read (sub { });
1577 $handle->on_eof (undef);
1578 $handle->on_error (sub {
1579 my $data = delete $_[0]{rbuf};
1580 undef $handle;
1581 });
1582
1583The reason to use C<on_error> is that TCP connections, due to latencies
1584and packets loss, might get closed quite violently with an error, when in
1585fact, all data has been received.
1586
1587It is usually better to use acknowledgements when transferring data,
1588to make sure the other side hasn't just died and you got the data
1589intact. This is also one reason why so many internet protocols have an
1590explicit QUIT command.
1591
1592=item I don't want to destroy the handle too early - how do I wait until
1593all data has been written?
1594
1595After writing your last bits of data, set the C<on_drain> callback
1596and destroy the handle in there - with the default setting of
1597C<low_water_mark> this will be called precisely when all data has been
1598written to the socket:
1599
1600 $handle->push_write (...);
1601 $handle->on_drain (sub {
1602 warn "all data submitted to the kernel\n";
1603 undef $handle;
1604 });
1605
1606=back
1607
1608
1609=head1 SUBCLASSING AnyEvent::Handle
1610
1611In many cases, you might want to subclass AnyEvent::Handle.
1612
1613To make this easier, a given version of AnyEvent::Handle uses these
1614conventions:
1615
1616=over 4
1617
1618=item * all constructor arguments become object members.
1619
1620At least initially, when you pass a C<tls>-argument to the constructor it
1621will end up in C<< $handle->{tls} >>. Those members might be changed or
1622mutated later on (for example C<tls> will hold the TLS connection object).
1623
1624=item * other object member names are prefixed with an C<_>.
1625
1626All object members not explicitly documented (internal use) are prefixed
1627with an underscore character, so the remaining non-C<_>-namespace is free
1628for use for subclasses.
1629
1630=item * all members not documented here and not prefixed with an underscore
1631are free to use in subclasses.
1632
1633Of course, new versions of AnyEvent::Handle may introduce more "public"
1634member variables, but thats just life, at least it is documented.
1635
1636=back
1637
798=head1 AUTHOR 1638=head1 AUTHOR
799 1639
800Robin Redeker C<< <elmex at ta-sa.org> >>, Marc Lehmann <schmorp@schmorp.de>. 1640Robin Redeker C<< <elmex at ta-sa.org> >>, Marc Lehmann <schmorp@schmorp.de>.
801 1641
802=cut 1642=cut

Diff Legend

Removed lines
+ Added lines
< Changed lines
> Changed lines