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 1.197 by root, Tue Aug 31 00:59:55 2010 UTC vs.
Revision 1.213 by root, Sat Jan 15 20:32:45 2011 UTC

… …
30 30
31 $cv->recv; 31 $cv->recv;
32 32
33=head1 DESCRIPTION 33=head1 DESCRIPTION
34 34
35This module is a helper module to make it easier to do event-based I/O on 35This is a helper module to make it easier to do event-based I/O on
36stream-based filehandles (sockets, pipes or other stream things). 36stream-based filehandles (sockets, pipes, and other stream things).
37 37
38The L<AnyEvent::Intro> tutorial contains some well-documented 38The L<AnyEvent::Intro> tutorial contains some well-documented
39AnyEvent::Handle examples. 39AnyEvent::Handle examples.
40 40
41In the following, when the documentation refers to of "bytes" then this 41In the following, where the documentation refers to "bytes", it means
42means characters. As sysread and syswrite are used for all I/O, their 42characters. As sysread and syswrite are used for all I/O, their
43treatment of characters applies to this module as well. 43treatment of characters applies to this module as well.
44 44
45At the very minimum, you should specify C<fh> or C<connect>, and the 45At the very minimum, you should specify C<fh> or C<connect>, and the
46C<on_error> callback. 46C<on_error> callback.
47 47
… …
75 } 75 }
76 76
77 \&$func 77 \&$func
78} 78}
79 79
80sub MAX_READ_SIZE() { 131072 }
81
80=head1 METHODS 82=head1 METHODS
81 83
82=over 4 84=over 4
83 85
84=item $handle = B<new> AnyEvent::Handle fh => $filehandle, key => value... 86=item $handle = B<new> AnyEvent::Handle fh => $filehandle, key => value...
… …
112=over 4 114=over 4
113 115
114=item on_prepare => $cb->($handle) 116=item on_prepare => $cb->($handle)
115 117
116This (rarely used) callback is called before a new connection is 118This (rarely used) callback is called before a new connection is
117attempted, but after the file handle has been created. It could be used to 119attempted, but after the file handle has been created (you can access that
120file handle via C<< $handle->{fh} >>). It could be used to prepare the
118prepare the file handle with parameters required for the actual connect 121file handle with parameters required for the actual connect (as opposed to
119(as opposed to settings that can be changed when the connection is already 122settings that can be changed when the connection is already established).
120established).
121 123
122The return value of this callback should be the connect timeout value in 124The return value of this callback should be the connect timeout value in
123seconds (or C<0>, or C<undef>, or the empty list, to indicate the default 125seconds (or C<0>, or C<undef>, or the empty list, to indicate that the
124timeout is to be used). 126default timeout is to be used).
125 127
126=item on_connect => $cb->($handle, $host, $port, $retry->()) 128=item on_connect => $cb->($handle, $host, $port, $retry->())
127 129
128This callback is called when a connection has been successfully established. 130This callback is called when a connection has been successfully established.
129 131
130The actual numeric host and port (the socket peername) are passed as 132The peer's numeric host and port (the socket peername) are passed as
131parameters, together with a retry callback. 133parameters, together with a retry callback.
132 134
133When, for some reason, the handle is not acceptable, then calling 135If, for some reason, the handle is not acceptable, calling C<$retry>
134C<$retry> will continue with the next connection target (in case of 136will continue with the next connection target (in case of multi-homed
135multi-homed hosts or SRV records there can be multiple connection 137hosts or SRV records there can be multiple connection endpoints). At the
136endpoints). At the time it is called the read and write queues, eof 138time it is called the read and write queues, eof status, tls status and
137status, tls status and similar properties of the handle will have been 139similar properties of the handle will have been reset.
138reset.
139 140
140In most cases, ignoring the C<$retry> parameter is the way to go. 141In most cases, you should ignore the C<$retry> parameter.
141 142
142=item on_connect_error => $cb->($handle, $message) 143=item on_connect_error => $cb->($handle, $message)
143 144
144This callback is called when the connection could not be 145This callback is called when the connection could not be
145established. C<$!> will contain the relevant error code, and C<$message> a 146established. C<$!> will contain the relevant error code, and C<$message> a
… …
152 153
153=item on_error => $cb->($handle, $fatal, $message) 154=item on_error => $cb->($handle, $fatal, $message)
154 155
155This is the error callback, which is called when, well, some error 156This is the error callback, which is called when, well, some error
156occured, such as not being able to resolve the hostname, failure to 157occured, such as not being able to resolve the hostname, failure to
157connect or a read error. 158connect, or a read error.
158 159
159Some errors are fatal (which is indicated by C<$fatal> being true). On 160Some errors are fatal (which is indicated by C<$fatal> being true). On
160fatal errors the handle object will be destroyed (by a call to C<< -> 161fatal errors the handle object will be destroyed (by a call to C<< ->
161destroy >>) after invoking the error callback (which means you are free to 162destroy >>) after invoking the error callback (which means you are free to
162examine the handle object). Examples of fatal errors are an EOF condition 163examine the handle object). Examples of fatal errors are an EOF condition
163with active (but unsatisifable) read watchers (C<EPIPE>) or I/O errors. In 164with active (but unsatisfiable) read watchers (C<EPIPE>) or I/O errors. In
164cases where the other side can close the connection at their will it is 165cases where the other side can close the connection at will, it is
165often easiest to not report C<EPIPE> errors in this callback. 166often easiest to not report C<EPIPE> errors in this callback.
166 167
167AnyEvent::Handle tries to find an appropriate error code for you to check 168AnyEvent::Handle tries to find an appropriate error code for you to check
168against, but in some cases (TLS errors), this does not work well. It is 169against, but in some cases (TLS errors), this does not work well. It is
169recommended to always output the C<$message> argument in human-readable 170recommended to always output the C<$message> argument in human-readable
170error messages (it's usually the same as C<"$!">). 171error messages (it's usually the same as C<"$!">).
171 172
172Non-fatal errors can be retried by simply returning, but it is recommended 173Non-fatal errors can be retried by returning, but it is recommended
173to simply ignore this parameter and instead abondon the handle object 174to simply ignore this parameter and instead abondon the handle object
174when this callback is invoked. Examples of non-fatal errors are timeouts 175when this callback is invoked. Examples of non-fatal errors are timeouts
175C<ETIMEDOUT>) or badly-formatted data (C<EBADMSG>). 176C<ETIMEDOUT>) or badly-formatted data (C<EBADMSG>).
176 177
177On callback entrance, the value of C<$!> contains the operating system 178On entry to the callback, the value of C<$!> contains the operating
178error code (or C<ENOSPC>, C<EPIPE>, C<ETIMEDOUT>, C<EBADMSG> or 179system error code (or C<ENOSPC>, C<EPIPE>, C<ETIMEDOUT>, C<EBADMSG> or
179C<EPROTO>). 180C<EPROTO>).
180 181
181While not mandatory, it is I<highly> recommended to set this callback, as 182While not mandatory, it is I<highly> recommended to set this callback, as
182you will not be notified of errors otherwise. The default simply calls 183you will not be notified of errors otherwise. The default just calls
183C<croak>. 184C<croak>.
184 185
185=item on_read => $cb->($handle) 186=item on_read => $cb->($handle)
186 187
187This sets the default read callback, which is called when data arrives 188This sets the default read callback, which is called when data arrives
… …
195the beginning from it. 196the beginning from it.
196 197
197You can also call C<< ->push_read (...) >> or any other function that 198You can also call C<< ->push_read (...) >> or any other function that
198modifies the read queue. Or do both. Or ... 199modifies the read queue. Or do both. Or ...
199 200
200When an EOF condition is detected then AnyEvent::Handle will first try to 201When an EOF condition is detected, AnyEvent::Handle will first try to
201feed all the remaining data to the queued callbacks and C<on_read> before 202feed all the remaining data to the queued callbacks and C<on_read> before
202calling the C<on_eof> callback. If no progress can be made, then a fatal 203calling the C<on_eof> callback. If no progress can be made, then a fatal
203error will be raised (with C<$!> set to C<EPIPE>). 204error will be raised (with C<$!> set to C<EPIPE>).
204 205
205Note that, unlike requests in the read queue, an C<on_read> callback 206Note that, unlike requests in the read queue, an C<on_read> callback
… …
224set, then a fatal error will be raised with C<$!> set to <0>. 225set, then a fatal error will be raised with C<$!> set to <0>.
225 226
226=item on_drain => $cb->($handle) 227=item on_drain => $cb->($handle)
227 228
228This sets the callback that is called when the write buffer becomes empty 229This sets the callback that is called when the write buffer becomes empty
229(or when the callback is set and the buffer is empty already). 230(or immediately if the buffer is empty already).
230 231
231To append to the write buffer, use the C<< ->push_write >> method. 232To append to the write buffer, use the C<< ->push_write >> method.
232 233
233This callback is useful when you don't want to put all of your write data 234This callback is useful when you don't want to put all of your write data
234into the queue at once, for example, when you want to write the contents 235into the queue at once, for example, when you want to write the contents
… …
246many seconds pass without a successful read or write on the underlying 247many seconds pass without a successful read or write on the underlying
247file handle (or a call to C<timeout_reset>), the C<on_timeout> callback 248file handle (or a call to C<timeout_reset>), the C<on_timeout> callback
248will be invoked (and if that one is missing, a non-fatal C<ETIMEDOUT> 249will be invoked (and if that one is missing, a non-fatal C<ETIMEDOUT>
249error will be raised). 250error will be raised).
250 251
251There are three variants of the timeouts that work fully independent 252There are three variants of the timeouts that work independently
252of each other, for both read and write, just read, and just write: 253of each other, for both read and write, just read, and just write:
253C<timeout>, C<rtimeout> and C<wtimeout>, with corresponding callbacks 254C<timeout>, C<rtimeout> and C<wtimeout>, with corresponding callbacks
254C<on_timeout>, C<on_rtimeout> and C<on_wtimeout>, and reset functions 255C<on_timeout>, C<on_rtimeout> and C<on_wtimeout>, and reset functions
255C<timeout_reset>, C<rtimeout_reset>, and C<wtimeout_reset>. 256C<timeout_reset>, C<rtimeout_reset>, and C<wtimeout_reset>.
256 257
257Note that timeout processing is also active when you currently do not have 258Note that timeout processing is active even when you do not have
258any outstanding read or write requests: If you plan to keep the connection 259any outstanding read or write requests: If you plan to keep the connection
259idle then you should disable the timout temporarily or ignore the timeout 260idle then you should disable the timeout temporarily or ignore the timeout
260in the C<on_timeout> callback, in which case AnyEvent::Handle will simply 261in the C<on_timeout> callback, in which case AnyEvent::Handle will simply
261restart the timeout. 262restart the timeout.
262 263
263Zero (the default) disables this timeout. 264Zero (the default) disables this timeout.
264 265
… …
278be configured to accept only so-and-so much data that it cannot act on 279be configured to accept only so-and-so much data that it cannot act on
279(for example, when expecting a line, an attacker could send an unlimited 280(for example, when expecting a line, an attacker could send an unlimited
280amount of data without a callback ever being called as long as the line 281amount of data without a callback ever being called as long as the line
281isn't finished). 282isn't finished).
282 283
284=item wbuf_max => <bytes>
285
286If defined, then a fatal error will be raised (with C<$!> set to C<ENOSPC>)
287when the write buffer ever (strictly) exceeds this size. This is useful to
288avoid some forms of denial-of-service attacks.
289
290Although the units of this parameter is bytes, this is the I<raw> number
291of bytes not yet accepted by the kernel. This can make a difference when
292you e.g. use TLS, as TLS typically makes your write data larger (but it
293can also make it smaller due to compression).
294
295As an example of when this limit is useful, take a chat server that sends
296chat messages to a client. If the client does not read those in a timely
297manner then the send buffer in the server would grow unbounded.
298
283=item autocork => <boolean> 299=item autocork => <boolean>
284 300
285When disabled (the default), then C<push_write> will try to immediately 301When disabled (the default), C<push_write> will try to immediately
286write the data to the handle, if possible. This avoids having to register 302write the data to the handle if possible. This avoids having to register
287a write watcher and wait for the next event loop iteration, but can 303a write watcher and wait for the next event loop iteration, but can
288be inefficient if you write multiple small chunks (on the wire, this 304be inefficient if you write multiple small chunks (on the wire, this
289disadvantage is usually avoided by your kernel's nagle algorithm, see 305disadvantage is usually avoided by your kernel's nagle algorithm, see
290C<no_delay>, but this option can save costly syscalls). 306C<no_delay>, but this option can save costly syscalls).
291 307
292When enabled, then writes will always be queued till the next event loop 308When enabled, writes will always be queued till the next event loop
293iteration. This is efficient when you do many small writes per iteration, 309iteration. This is efficient when you do many small writes per iteration,
294but less efficient when you do a single write only per iteration (or when 310but less efficient when you do a single write only per iteration (or when
295the write buffer often is full). It also increases write latency. 311the write buffer often is full). It also increases write latency.
296 312
297=item no_delay => <boolean> 313=item no_delay => <boolean>
… …
301the Nagle algorithm, and usually it is beneficial. 317the Nagle algorithm, and usually it is beneficial.
302 318
303In some situations you want as low a delay as possible, which can be 319In some situations you want as low a delay as possible, which can be
304accomplishd by setting this option to a true value. 320accomplishd by setting this option to a true value.
305 321
306The default is your opertaing system's default behaviour (most likely 322The default is your operating system's default behaviour (most likely
307enabled), this option explicitly enables or disables it, if possible. 323enabled). This option explicitly enables or disables it, if possible.
308 324
309=item keepalive => <boolean> 325=item keepalive => <boolean>
310 326
311Enables (default disable) the SO_KEEPALIVE option on the stream socket: 327Enables (default disable) the SO_KEEPALIVE option on the stream socket:
312normally, TCP connections have no time-out once established, so TCP 328normally, TCP connections have no time-out once established, so TCP
313connections, once established, can stay alive forever even when the other 329connections, once established, can stay alive forever even when the other
314side has long gone. TCP keepalives are a cheap way to take down long-lived 330side has long gone. TCP keepalives are a cheap way to take down long-lived
315TCP connections whent he other side becomes unreachable. While the default 331TCP connections when the other side becomes unreachable. While the default
316is OS-dependent, TCP keepalives usually kick in after around two hours, 332is OS-dependent, TCP keepalives usually kick in after around two hours,
317and, if the other side doesn't reply, take down the TCP connection some 10 333and, if the other side doesn't reply, take down the TCP connection some 10
318to 15 minutes later. 334to 15 minutes later.
319 335
320It is harmless to specify this option for file handles that do not support 336It is harmless to specify this option for file handles that do not support
… …
338already have occured on BSD systems), but at least it will protect you 354already have occured on BSD systems), but at least it will protect you
339from most attacks. 355from most attacks.
340 356
341=item read_size => <bytes> 357=item read_size => <bytes>
342 358
343The default read block size (the amount of bytes this module will 359The initial read block size, the number of bytes this module will try to
344try to read during each loop iteration, which affects memory 360read during each loop iteration. Each handle object will consume at least
345requirements). Default: C<8192>. 361this amount of memory for the read buffer as well, so when handling many
362connections requirements). See also C<max_read_size>. Default: C<2048>.
363
364=item max_read_size => <bytes>
365
366The maximum read buffer size used by the dynamic adjustment
367algorithm: Each time AnyEvent::Handle can read C<read_size> bytes in
368one go it will double C<read_size> up to the maximum given by this
369option. Default: C<131072> or C<read_size>, whichever is higher.
346 370
347=item low_water_mark => <bytes> 371=item low_water_mark => <bytes>
348 372
349Sets the amount of bytes (default: C<0>) that make up an "empty" write 373Sets the number of bytes (default: C<0>) that make up an "empty" write
350buffer: If the write reaches this size or gets even samller it is 374buffer: If the buffer reaches this size or gets even samller it is
351considered empty. 375considered empty.
352 376
353Sometimes it can be beneficial (for performance reasons) to add data to 377Sometimes it can be beneficial (for performance reasons) to add data to
354the write buffer before it is fully drained, but this is a rare case, as 378the write buffer before it is fully drained, but this is a rare case, as
355the operating system kernel usually buffers data as well, so the default 379the operating system kernel usually buffers data as well, so the default
356is good in almost all cases. 380is good in almost all cases.
357 381
358=item linger => <seconds> 382=item linger => <seconds>
359 383
360If non-zero (default: C<3600>), then the destructor of the 384If this is non-zero (default: C<3600>), the destructor of the
361AnyEvent::Handle object will check whether there is still outstanding 385AnyEvent::Handle object will check whether there is still outstanding
362write data and will install a watcher that will write this data to the 386write data and will install a watcher that will write this data to the
363socket. No errors will be reported (this mostly matches how the operating 387socket. No errors will be reported (this mostly matches how the operating
364system treats outstanding data at socket close time). 388system treats outstanding data at socket close time).
365 389
… …
372A string used to identify the remote site - usually the DNS hostname 396A string used to identify the remote site - usually the DNS hostname
373(I<not> IDN!) used to create the connection, rarely the IP address. 397(I<not> IDN!) used to create the connection, rarely the IP address.
374 398
375Apart from being useful in error messages, this string is also used in TLS 399Apart from being useful in error messages, this string is also used in TLS
376peername verification (see C<verify_peername> in L<AnyEvent::TLS>). This 400peername verification (see C<verify_peername> in L<AnyEvent::TLS>). This
377verification will be skipped when C<peername> is not specified or 401verification will be skipped when C<peername> is not specified or is
378C<undef>. 402C<undef>.
379 403
380=item tls => "accept" | "connect" | Net::SSLeay::SSL object 404=item tls => "accept" | "connect" | Net::SSLeay::SSL object
381 405
382When this parameter is given, it enables TLS (SSL) mode, that means 406When this parameter is given, it enables TLS (SSL) mode, that means
… …
408B<IMPORTANT:> since Net::SSLeay "objects" are really only integers, 432B<IMPORTANT:> since Net::SSLeay "objects" are really only integers,
409passing in the wrong integer will lead to certain crash. This most often 433passing in the wrong integer will lead to certain crash. This most often
410happens when one uses a stylish C<< tls => 1 >> and is surprised about the 434happens when one uses a stylish C<< tls => 1 >> and is surprised about the
411segmentation fault. 435segmentation fault.
412 436
413See the C<< ->starttls >> method for when need to start TLS negotiation later. 437Use the C<< ->starttls >> method if you need to start TLS negotiation later.
414 438
415=item tls_ctx => $anyevent_tls 439=item tls_ctx => $anyevent_tls
416 440
417Use the given C<AnyEvent::TLS> object to create the new TLS connection 441Use the given C<AnyEvent::TLS> object to create the new TLS connection
418(unless a connection object was specified directly). If this parameter is 442(unless a connection object was specified directly). If this
419missing, then AnyEvent::Handle will use C<AnyEvent::Handle::TLS_CTX>. 443parameter is missing (or C<undef>), then AnyEvent::Handle will use
444C<AnyEvent::Handle::TLS_CTX>.
420 445
421Instead of an object, you can also specify a hash reference with C<< key 446Instead of an object, you can also specify a hash reference with C<< key
422=> value >> pairs. Those will be passed to L<AnyEvent::TLS> to create a 447=> value >> pairs. Those will be passed to L<AnyEvent::TLS> to create a
423new TLS context object. 448new TLS context object.
424 449
… …
433 458
434TLS handshake failures will not cause C<on_error> to be invoked when this 459TLS handshake failures will not cause C<on_error> to be invoked when this
435callback is in effect, instead, the error message will be passed to C<on_starttls>. 460callback is in effect, instead, the error message will be passed to C<on_starttls>.
436 461
437Without this callback, handshake failures lead to C<on_error> being 462Without this callback, handshake failures lead to C<on_error> being
438called, as normal. 463called as usual.
439 464
440Note that you cannot call C<starttls> right again in this callback. If you 465Note that you cannot just call C<starttls> again in this callback. If you
441need to do that, start an zero-second timer instead whose callback can 466need to do that, start an zero-second timer instead whose callback can
442then call C<< ->starttls >> again. 467then call C<< ->starttls >> again.
443 468
444=item on_stoptls => $cb->($handle) 469=item on_stoptls => $cb->($handle)
445 470
… …
492 AnyEvent::Socket::tcp_connect ( 517 AnyEvent::Socket::tcp_connect (
493 $self->{connect}[0], 518 $self->{connect}[0],
494 $self->{connect}[1], 519 $self->{connect}[1],
495 sub { 520 sub {
496 my ($fh, $host, $port, $retry) = @_; 521 my ($fh, $host, $port, $retry) = @_;
522
523 delete $self->{_connect}; # no longer needed
497 524
498 if ($fh) { 525 if ($fh) {
499 $self->{fh} = $fh; 526 $self->{fh} = $fh;
500 527
501 delete $self->{_skip_drain_rbuf}; 528 delete $self->{_skip_drain_rbuf};
… …
519 }, 546 },
520 sub { 547 sub {
521 local $self->{fh} = $_[0]; 548 local $self->{fh} = $_[0];
522 549
523 $self->{on_prepare} 550 $self->{on_prepare}
524 ? $self->{on_prepare}->($self) 551 ? $self->{on_prepare}->($self)
525 : () 552 : ()
526 } 553 }
527 ); 554 );
528 } 555 }
529 556
… …
547 574
548 $self->{_activity} = 575 $self->{_activity} =
549 $self->{_ractivity} = 576 $self->{_ractivity} =
550 $self->{_wactivity} = AE::now; 577 $self->{_wactivity} = AE::now;
551 578
579 $self->{read_size} ||= 2048;
580 $self->{max_read_size} = $self->{read_size}
581 if $self->{read_size} > ($self->{max_read_size} || MAX_READ_SIZE);
582
552 $self->timeout (delete $self->{timeout} ) if $self->{timeout}; 583 $self->timeout (delete $self->{timeout} ) if $self->{timeout};
553 $self->rtimeout (delete $self->{rtimeout} ) if $self->{rtimeout}; 584 $self->rtimeout (delete $self->{rtimeout} ) if $self->{rtimeout};
554 $self->wtimeout (delete $self->{wtimeout} ) if $self->{wtimeout}; 585 $self->wtimeout (delete $self->{wtimeout} ) if $self->{wtimeout};
555 586
556 $self->no_delay (delete $self->{no_delay} ) if exists $self->{no_delay} && $self->{no_delay}; 587 $self->no_delay (delete $self->{no_delay} ) if exists $self->{no_delay} && $self->{no_delay};
… …
559 $self->oobinline (exists $self->{oobinline} ? delete $self->{oobinline} : 1); 590 $self->oobinline (exists $self->{oobinline} ? delete $self->{oobinline} : 1);
560 591
561 $self->starttls (delete $self->{tls}, delete $self->{tls_ctx}) 592 $self->starttls (delete $self->{tls}, delete $self->{tls_ctx})
562 if $self->{tls}; 593 if $self->{tls};
563 594
564 $self->on_drain (delete $self->{on_drain}) if $self->{on_drain}; 595 $self->on_drain (delete $self->{on_drain} ) if $self->{on_drain};
565 596
566 $self->start_read 597 $self->start_read
567 if $self->{on_read} || @{ $self->{_queue} }; 598 if $self->{on_read} || @{ $self->{_queue} };
568 599
569 $self->_drain_wbuf; 600 $self->_drain_wbuf;
… …
645=cut 676=cut
646 677
647sub no_delay { 678sub no_delay {
648 $_[0]{no_delay} = $_[1]; 679 $_[0]{no_delay} = $_[1];
649 680
650 eval {
651 local $SIG{__DIE__};
652 setsockopt $_[0]{fh}, Socket::IPPROTO_TCP (), Socket::TCP_NODELAY (), int $_[1] 681 setsockopt $_[0]{fh}, Socket::IPPROTO_TCP (), Socket::TCP_NODELAY (), int $_[1]
653 if $_[0]{fh}; 682 if $_[0]{fh};
654 };
655} 683}
656 684
657=item $handle->keepalive ($boolean) 685=item $handle->keepalive ($boolean)
658 686
659Enables or disables the C<keepalive> setting (see constructor argument of 687Enables or disables the C<keepalive> setting (see constructor argument of
… …
727 755
728=item $handle->rbuf_max ($max_octets) 756=item $handle->rbuf_max ($max_octets)
729 757
730Configures the C<rbuf_max> setting (C<undef> disables it). 758Configures the C<rbuf_max> setting (C<undef> disables it).
731 759
760=item $handle->wbuf_max ($max_octets)
761
762Configures the C<wbuf_max> setting (C<undef> disables it).
763
732=cut 764=cut
733 765
734sub rbuf_max { 766sub rbuf_max {
735 $_[0]{rbuf_max} = $_[1]; 767 $_[0]{rbuf_max} = $_[1];
768}
769
770sub rbuf_max {
771 $_[0]{wbuf_max} = $_[1];
736} 772}
737 773
738############################################################################# 774#############################################################################
739 775
740=item $handle->timeout ($seconds) 776=item $handle->timeout ($seconds)
… …
768 $_[0]{$on_timeout} = $_[1]; 804 $_[0]{$on_timeout} = $_[1];
769 }; 805 };
770 806
771 *$timeout = sub { 807 *$timeout = sub {
772 my ($self, $new_value) = @_; 808 my ($self, $new_value) = @_;
809
810 $new_value >= 0
811 or Carp::croak "AnyEvent::Handle->$timeout called with negative timeout ($new_value), caught";
773 812
774 $self->{$timeout} = $new_value; 813 $self->{$timeout} = $new_value;
775 delete $self->{$tw}; &$cb; 814 delete $self->{$tw}; &$cb;
776 }; 815 };
777 816
… …
857 if $cb && $self->{low_water_mark} >= (length $self->{wbuf}) + (length $self->{_tls_wbuf}); 896 if $cb && $self->{low_water_mark} >= (length $self->{wbuf}) + (length $self->{_tls_wbuf});
858} 897}
859 898
860=item $handle->push_write ($data) 899=item $handle->push_write ($data)
861 900
862Queues the given scalar to be written. You can push as much data as you 901Queues the given scalar to be written. You can push as much data as
863want (only limited by the available memory), as C<AnyEvent::Handle> 902you want (only limited by the available memory and C<wbuf_max>), as
864buffers it independently of the kernel. 903C<AnyEvent::Handle> buffers it independently of the kernel.
865 904
866This method may invoke callbacks (and therefore the handle might be 905This method may invoke callbacks (and therefore the handle might be
867destroyed after it returns). 906destroyed after it returns).
868 907
869=cut 908=cut
… …
897 $cb->() unless $self->{autocork}; 936 $cb->() unless $self->{autocork};
898 937
899 # if still data left in wbuf, we need to poll 938 # if still data left in wbuf, we need to poll
900 $self->{_ww} = AE::io $self->{fh}, 1, $cb 939 $self->{_ww} = AE::io $self->{fh}, 1, $cb
901 if length $self->{wbuf}; 940 if length $self->{wbuf};
941
942 if (
943 defined $self->{wbuf_max}
944 && $self->{wbuf_max} < length $self->{wbuf}
945 ) {
946 $self->_error (Errno::ENOSPC, 1), return;
947 }
902 }; 948 };
903} 949}
904 950
905our %WH; 951our %WH;
906 952
… …
1117partial message has been received so far), or change the read queue with 1163partial message has been received so far), or change the read queue with
1118e.g. C<push_read>. 1164e.g. C<push_read>.
1119 1165
1120In the more complex case, you want to queue multiple callbacks. In this 1166In the more complex case, you want to queue multiple callbacks. In this
1121case, AnyEvent::Handle will call the first queued callback each time new 1167case, AnyEvent::Handle will call the first queued callback each time new
1122data arrives (also the first time it is queued) and removes it when it has 1168data arrives (also the first time it is queued) and remove it when it has
1123done its job (see C<push_read>, below). 1169done its job (see C<push_read>, below).
1124 1170
1125This way you can, for example, push three line-reads, followed by reading 1171This way you can, for example, push three line-reads, followed by reading
1126a chunk of data, and AnyEvent::Handle will execute them in order. 1172a chunk of data, and AnyEvent::Handle will execute them in order.
1127 1173
… …
1272 $self->_drain_rbuf if $cb; 1318 $self->_drain_rbuf if $cb;
1273} 1319}
1274 1320
1275=item $handle->rbuf 1321=item $handle->rbuf
1276 1322
1277Returns the read buffer (as a modifiable lvalue). 1323Returns the read buffer (as a modifiable lvalue). You can also access the
1324read buffer directly as the C<< ->{rbuf} >> member, if you want (this is
1325much faster, and no less clean).
1278 1326
1279You can access the read buffer directly as the C<< ->{rbuf} >> 1327The only operation allowed on the read buffer (apart from looking at it)
1280member, if you want. However, the only operation allowed on the 1328is removing data from its beginning. Otherwise modifying or appending to
1281read buffer (apart from looking at it) is removing data from its 1329it is not allowed and will lead to hard-to-track-down bugs.
1282beginning. Otherwise modifying or appending to it is not allowed and will
1283lead to hard-to-track-down bugs.
1284 1330
1285NOTE: The read buffer should only be used or modified if the C<on_read>, 1331NOTE: The read buffer should only be used or modified in the C<on_read>
1286C<push_read> or C<unshift_read> methods are used. The other read methods 1332callback or when C<push_read> or C<unshift_read> are used with a single
1287automatically manage the read buffer. 1333callback (i.e. untyped). Typed C<push_read> and C<unshift_read> methods
1334will manage the read buffer on their own.
1288 1335
1289=cut 1336=cut
1290 1337
1291sub rbuf : lvalue { 1338sub rbuf : lvalue {
1292 $_[0]{rbuf} 1339 $_[0]{rbuf}
… …
1343 my $cb = pop; 1390 my $cb = pop;
1344 1391
1345 if (@_) { 1392 if (@_) {
1346 my $type = shift; 1393 my $type = shift;
1347 1394
1395 $cb = ($RH{$type} ||= _load_func "$type\::anyevent_read_type"
1348 $cb = ($RH{$type} or Carp::croak "unsupported type passed to AnyEvent::Handle::unshift_read") 1396 or Carp::croak "unsupported/unloadable type '$type' passed to AnyEvent::Handle::unshift_read")
1349 ->($self, $cb, @_); 1397 ->($self, $cb, @_);
1350 } 1398 }
1351 1399
1352 unshift @{ $self->{_queue} }, $cb; 1400 unshift @{ $self->{_queue} }, $cb;
1353 $self->_drain_rbuf; 1401 $self->_drain_rbuf;
… …
1461the receive buffer when neither C<$accept> nor C<$reject> match, 1509the receive buffer when neither C<$accept> nor C<$reject> match,
1462and everything preceding and including the match will be accepted 1510and everything preceding and including the match will be accepted
1463unconditionally. This is useful to skip large amounts of data that you 1511unconditionally. This is useful to skip large amounts of data that you
1464know cannot be matched, so that the C<$accept> or C<$reject> regex do not 1512know cannot be matched, so that the C<$accept> or C<$reject> regex do not
1465have to start matching from the beginning. This is purely an optimisation 1513have to start matching from the beginning. This is purely an optimisation
1466and is usually worth only when you expect more than a few kilobytes. 1514and is usually worth it only when you expect more than a few kilobytes.
1467 1515
1468Example: expect a http header, which ends at C<\015\012\015\012>. Since we 1516Example: expect a http header, which ends at C<\015\012\015\012>. Since we
1469expect the header to be very large (it isn't in practise, but...), we use 1517expect the header to be very large (it isn't in practice, but...), we use
1470a skip regex to skip initial portions. The skip regex is tricky in that 1518a skip regex to skip initial portions. The skip regex is tricky in that
1471it only accepts something not ending in either \015 or \012, as these are 1519it only accepts something not ending in either \015 or \012, as these are
1472required for the accept regex. 1520required for the accept regex.
1473 1521
1474 $handle->push_read (regex => 1522 $handle->push_read (regex =>
… …
1725Note that AnyEvent::Handle will automatically C<start_read> for you when 1773Note that AnyEvent::Handle will automatically C<start_read> for you when
1726you change the C<on_read> callback or push/unshift a read callback, and it 1774you change the C<on_read> callback or push/unshift a read callback, and it
1727will automatically C<stop_read> for you when neither C<on_read> is set nor 1775will automatically C<stop_read> for you when neither C<on_read> is set nor
1728there are any read requests in the queue. 1776there are any read requests in the queue.
1729 1777
1730These methods will have no effect when in TLS mode (as TLS doesn't support 1778In older versions of this module (<= 5.3), these methods had no effect,
1731half-duplex connections). 1779as TLS does not support half-duplex connections. In current versions they
1780work as expected, as this behaviour is required to avoid certain resource
1781attacks, where the program would be forced to read (and buffer) arbitrary
1782amounts of data before being able to send some data. The drawback is that
1783some readings of the the SSL/TLS specifications basically require this
1784attack to be working, as SSL/TLS implementations might stall sending data
1785during a rehandshake.
1786
1787As a guideline, during the initial handshake, you should not stop reading,
1788and as a client, it might cause problems, depending on your applciation.
1732 1789
1733=cut 1790=cut
1734 1791
1735sub stop_read { 1792sub stop_read {
1736 my ($self) = @_; 1793 my ($self) = @_;
1737 1794
1738 delete $self->{_rw} unless $self->{tls}; 1795 delete $self->{_rw};
1739} 1796}
1740 1797
1741sub start_read { 1798sub start_read {
1742 my ($self) = @_; 1799 my ($self) = @_;
1743 1800
1744 unless ($self->{_rw} || $self->{_eof} || !$self->{fh}) { 1801 unless ($self->{_rw} || $self->{_eof} || !$self->{fh}) {
1745 Scalar::Util::weaken $self; 1802 Scalar::Util::weaken $self;
1746 1803
1747 $self->{_rw} = AE::io $self->{fh}, 0, sub { 1804 $self->{_rw} = AE::io $self->{fh}, 0, sub {
1748 my $rbuf = \($self->{tls} ? my $buf : $self->{rbuf}); 1805 my $rbuf = \($self->{tls} ? my $buf : $self->{rbuf});
1749 my $len = sysread $self->{fh}, $$rbuf, $self->{read_size} || 8192, length $$rbuf; 1806 my $len = sysread $self->{fh}, $$rbuf, $self->{read_size}, length $$rbuf;
1750 1807
1751 if ($len > 0) { 1808 if ($len > 0) {
1752 $self->{_activity} = $self->{_ractivity} = AE::now; 1809 $self->{_activity} = $self->{_ractivity} = AE::now;
1753 1810
1754 if ($self->{tls}) { 1811 if ($self->{tls}) {
1755 Net::SSLeay::BIO_write ($self->{_rbio}, $$rbuf); 1812 Net::SSLeay::BIO_write ($self->{_rbio}, $$rbuf);
1756 1813
1757 &_dotls ($self); 1814 &_dotls ($self);
1758 } else { 1815 } else {
1759 $self->_drain_rbuf; 1816 $self->_drain_rbuf;
1817 }
1818
1819 if ($len == $self->{read_size}) {
1820 $self->{read_size} *= 2;
1821 $self->{read_size} = $self->{max_read_size} || MAX_READ_SIZE
1822 if $self->{read_size} > ($self->{max_read_size} || MAX_READ_SIZE);
1760 } 1823 }
1761 1824
1762 } elsif (defined $len) { 1825 } elsif (defined $len) {
1763 delete $self->{_rw}; 1826 delete $self->{_rw};
1764 $self->{_eof} = 1; 1827 $self->{_eof} = 1;
… …
1874context in C<< $handle->{tls_ctx} >> after this call and can be used or 1937context in C<< $handle->{tls_ctx} >> after this call and can be used or
1875changed to your liking. Note that the handshake might have already started 1938changed to your liking. Note that the handshake might have already started
1876when this function returns. 1939when this function returns.
1877 1940
1878Due to bugs in OpenSSL, it might or might not be possible to do multiple 1941Due to bugs in OpenSSL, it might or might not be possible to do multiple
1879handshakes on the same stream. Best do not attempt to use the stream after 1942handshakes on the same stream. It is best to not attempt to use the
1880stopping TLS. 1943stream after stopping TLS.
1881 1944
1882This method may invoke callbacks (and therefore the handle might be 1945This method may invoke callbacks (and therefore the handle might be
1883destroyed after it returns). 1946destroyed after it returns).
1884 1947
1885=cut 1948=cut
… …
2005 push @linger, AE::io $fh, 1, sub { 2068 push @linger, AE::io $fh, 1, sub {
2006 my $len = syswrite $fh, $wbuf, length $wbuf; 2069 my $len = syswrite $fh, $wbuf, length $wbuf;
2007 2070
2008 if ($len > 0) { 2071 if ($len > 0) {
2009 substr $wbuf, 0, $len, ""; 2072 substr $wbuf, 0, $len, "";
2010 } else { 2073 } elsif (defined $len || ($! != EAGAIN && $! != EINTR && $! != WSAEWOULDBLOCK)) {
2011 @linger = (); # end 2074 @linger = (); # end
2012 } 2075 }
2013 }; 2076 };
2014 push @linger, AE::timer $linger, 0, sub { 2077 push @linger, AE::timer $linger, 0, sub {
2015 @linger = (); 2078 @linger = ();
… …
2111 2174
2112It is only safe to "forget" the reference inside EOF or error callbacks, 2175It is only safe to "forget" the reference inside EOF or error callbacks,
2113from within all other callbacks, you need to explicitly call the C<< 2176from within all other callbacks, you need to explicitly call the C<<
2114->destroy >> method. 2177->destroy >> method.
2115 2178
2179=item Why is my C<on_eof> callback never called?
2180
2181Probably because your C<on_error> callback is being called instead: When
2182you have outstanding requests in your read queue, then an EOF is
2183considered an error as you clearly expected some data.
2184
2185To avoid this, make sure you have an empty read queue whenever your handle
2186is supposed to be "idle" (i.e. connection closes are O.K.). You cna set
2187an C<on_read> handler that simply pushes the first read requests in the
2188queue.
2189
2190See also the next question, which explains this in a bit more detail.
2191
2192=item How can I serve requests in a loop?
2193
2194Most protocols consist of some setup phase (authentication for example)
2195followed by a request handling phase, where the server waits for requests
2196and handles them, in a loop.
2197
2198There are two important variants: The first (traditional, better) variant
2199handles requests until the server gets some QUIT command, causing it to
2200close the connection first (highly desirable for a busy TCP server). A
2201client dropping the connection is an error, which means this variant can
2202detect an unexpected detection close.
2203
2204To handle this case, always make sure you have a on-empty read queue, by
2205pushing the "read request start" handler on it:
2206
2207 # we assume a request starts with a single line
2208 my @start_request; @start_request = (line => sub {
2209 my ($hdl, $line) = @_;
2210
2211 ... handle request
2212
2213 # push next request read, possibly from a nested callback
2214 $hdl->push_read (@start_request);
2215 });
2216
2217 # auth done, now go into request handling loop
2218 # now push the first @start_request
2219 $hdl->push_read (@start_request);
2220
2221By always having an outstanding C<push_read>, the handle always expects
2222some data and raises the C<EPIPE> error when the connction is dropped
2223unexpectedly.
2224
2225The second variant is a protocol where the client can drop the connection
2226at any time. For TCP, this means that the server machine may run out of
2227sockets easier, and in general, it means you cnanot distinguish a protocl
2228failure/client crash from a normal connection close. Nevertheless, these
2229kinds of protocols are common (and sometimes even the best solution to the
2230problem).
2231
2232Having an outstanding read request at all times is possible if you ignore
2233C<EPIPE> errors, but this doesn't help with when the client drops the
2234connection during a request, which would still be an error.
2235
2236A better solution is to push the initial request read in an C<on_read>
2237callback. This avoids an error, as when the server doesn't expect data
2238(i.e. is idly waiting for the next request, an EOF will not raise an
2239error, but simply result in an C<on_eof> callback. It is also a bit slower
2240and simpler:
2241
2242 # auth done, now go into request handling loop
2243 $hdl->on_read (sub {
2244 my ($hdl) = @_;
2245
2246 # called each time we receive data but the read queue is empty
2247 # simply start read the request
2248
2249 $hdl->push_read (line => sub {
2250 my ($hdl, $line) = @_;
2251
2252 ... handle request
2253
2254 # do nothing special when the request has been handled, just
2255 # let the request queue go empty.
2256 });
2257 });
2258
2116=item I get different callback invocations in TLS mode/Why can't I pause 2259=item I get different callback invocations in TLS mode/Why can't I pause
2117reading? 2260reading?
2118 2261
2119Unlike, say, TCP, TLS connections do not consist of two independent 2262Unlike, say, TCP, TLS connections do not consist of two independent
2120communication channels, one for each direction. Or put differently. The 2263communication channels, one for each direction. Or put differently, the
2121read and write directions are not independent of each other: you cannot 2264read and write directions are not independent of each other: you cannot
2122write data unless you are also prepared to read, and vice versa. 2265write data unless you are also prepared to read, and vice versa.
2123 2266
2124This can mean than, in TLS mode, you might get C<on_error> or C<on_eof> 2267This means that, in TLS mode, you might get C<on_error> or C<on_eof>
2125callback invocations when you are not expecting any read data - the reason 2268callback invocations when you are not expecting any read data - the reason
2126is that AnyEvent::Handle always reads in TLS mode. 2269is that AnyEvent::Handle always reads in TLS mode.
2127 2270
2128During the connection, you have to make sure that you always have a 2271During the connection, you have to make sure that you always have a
2129non-empty read-queue, or an C<on_read> watcher. At the end of the 2272non-empty read-queue, or an C<on_read> watcher. At the end of the
… …
2143 my $data = delete $_[0]{rbuf}; 2286 my $data = delete $_[0]{rbuf};
2144 }); 2287 });
2145 2288
2146The reason to use C<on_error> is that TCP connections, due to latencies 2289The reason to use C<on_error> is that TCP connections, due to latencies
2147and packets loss, might get closed quite violently with an error, when in 2290and packets loss, might get closed quite violently with an error, when in
2148fact, all data has been received. 2291fact all data has been received.
2149 2292
2150It is usually better to use acknowledgements when transferring data, 2293It is usually better to use acknowledgements when transferring data,
2151to make sure the other side hasn't just died and you got the data 2294to make sure the other side hasn't just died and you got the data
2152intact. This is also one reason why so many internet protocols have an 2295intact. This is also one reason why so many internet protocols have an
2153explicit QUIT command. 2296explicit QUIT command.
… …
2170consider using C<< ->push_shutdown >> instead. 2313consider using C<< ->push_shutdown >> instead.
2171 2314
2172=item I want to contact a TLS/SSL server, I don't care about security. 2315=item I want to contact a TLS/SSL server, I don't care about security.
2173 2316
2174If your TLS server is a pure TLS server (e.g. HTTPS) that only speaks TLS, 2317If your TLS server is a pure TLS server (e.g. HTTPS) that only speaks TLS,
2175simply connect to it and then create the AnyEvent::Handle with the C<tls> 2318connect to it and then create the AnyEvent::Handle with the C<tls>
2176parameter: 2319parameter:
2177 2320
2178 tcp_connect $host, $port, sub { 2321 tcp_connect $host, $port, sub {
2179 my ($fh) = @_; 2322 my ($fh) = @_;
2180 2323
… …
2280 2423
2281=item * all members not documented here and not prefixed with an underscore 2424=item * all members not documented here and not prefixed with an underscore
2282are free to use in subclasses. 2425are free to use in subclasses.
2283 2426
2284Of course, new versions of AnyEvent::Handle may introduce more "public" 2427Of course, new versions of AnyEvent::Handle may introduce more "public"
2285member variables, but thats just life, at least it is documented. 2428member variables, but that's just life. At least it is documented.
2286 2429
2287=back 2430=back
2288 2431
2289=head1 AUTHOR 2432=head1 AUTHOR
2290 2433

Diff Legend

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