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 1.177 by root, Sun Aug 9 00:24:35 2009 UTC vs.
Revision 1.218 by root, Thu Feb 24 12:04:20 2011 UTC

1=head1 NAME 1=head1 NAME
2 2
3AnyEvent::Handle - non-blocking I/O on file handles via AnyEvent 3AnyEvent::Handle - non-blocking I/O on streaming handles via AnyEvent
4 4
5=head1 SYNOPSIS 5=head1 SYNOPSIS
6 6
7 use AnyEvent; 7 use AnyEvent;
8 use AnyEvent::Handle; 8 use AnyEvent::Handle;
… …
14 on_error => sub { 14 on_error => sub {
15 my ($hdl, $fatal, $msg) = @_; 15 my ($hdl, $fatal, $msg) = @_;
16 warn "got error $msg\n"; 16 warn "got error $msg\n";
17 $hdl->destroy; 17 $hdl->destroy;
18 $cv->send; 18 $cv->send;
19 ); 19 };
20 20
21 # send some request line 21 # send some request line
22 $hdl->push_write ("getinfo\015\012"); 22 $hdl->push_write ("getinfo\015\012");
23 23
24 # read the response line 24 # read the response line
… …
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
36filehandles. 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
… …
60use AnyEvent (); BEGIN { AnyEvent::common_sense } 60use AnyEvent (); BEGIN { AnyEvent::common_sense }
61use AnyEvent::Util qw(WSAEWOULDBLOCK); 61use AnyEvent::Util qw(WSAEWOULDBLOCK);
62 62
63our $VERSION = $AnyEvent::VERSION; 63our $VERSION = $AnyEvent::VERSION;
64 64
65sub _load_func($) {
66 my $func = $_[0];
67
68 unless (defined &$func) {
69 my $pkg = $func;
70 do {
71 $pkg =~ s/::[^:]+$//
72 or return;
73 eval "require $pkg";
74 } until defined &$func;
75 }
76
77 \&$func
78}
79
80sub MAX_READ_SIZE() { 131072 }
81
65=head1 METHODS 82=head1 METHODS
66 83
67=over 4 84=over 4
68 85
69=item $handle = B<new> AnyEvent::TLS fh => $filehandle, key => value... 86=item $handle = B<new> AnyEvent::Handle fh => $filehandle, key => value...
70 87
71The constructor supports these arguments (all as C<< key => value >> pairs). 88The constructor supports these arguments (all as C<< key => value >> pairs).
72 89
73=over 4 90=over 4
74 91
… …
97=over 4 114=over 4
98 115
99=item on_prepare => $cb->($handle) 116=item on_prepare => $cb->($handle)
100 117
101This (rarely used) callback is called before a new connection is 118This (rarely used) callback is called before a new connection is
102attempted, 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
103prepare the file handle with parameters required for the actual connect 121file handle with parameters required for the actual connect (as opposed to
104(as opposed to settings that can be changed when the connection is already 122settings that can be changed when the connection is already established).
105established).
106 123
107The 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
108seconds (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
109timeout is to be used). 126default timeout is to be used).
110 127
111=item on_connect => $cb->($handle, $host, $port, $retry->()) 128=item on_connect => $cb->($handle, $host, $port, $retry->())
112 129
113This callback is called when a connection has been successfully established. 130This callback is called when a connection has been successfully established.
114 131
115The actual numeric host and port (the socket peername) are passed as 132The peer's numeric host and port (the socket peername) are passed as
116parameters, together with a retry callback. 133parameters, together with a retry callback.
117 134
118When, for some reason, the handle is not acceptable, then calling 135If, for some reason, the handle is not acceptable, calling C<$retry>
119C<$retry> will continue with the next conenction target (in case of 136will continue with the next connection target (in case of multi-homed
120multi-homed hosts or SRV records there can be multiple connection 137hosts or SRV records there can be multiple connection endpoints). At the
121endpoints). When it is called then the read and write queues, eof status, 138time it is called the read and write queues, eof status, tls status and
122tls status and similar properties of the handle are being reset. 139similar properties of the handle will have been reset.
123 140
124In most cases, ignoring the C<$retry> parameter is the way to go. 141In most cases, you should ignore the C<$retry> parameter.
125 142
126=item on_connect_error => $cb->($handle, $message) 143=item on_connect_error => $cb->($handle, $message)
127 144
128This callback is called when the conenction could not be 145This callback is called when the connection could not be
129established. C<$!> will contain the relevant error code, and C<$message> a 146established. C<$!> will contain the relevant error code, and C<$message> a
130message describing it (usually the same as C<"$!">). 147message describing it (usually the same as C<"$!">).
131 148
132If this callback isn't specified, then C<on_error> will be called with a 149If this callback isn't specified, then C<on_error> will be called with a
133fatal error instead. 150fatal error instead.
… …
136 153
137=item on_error => $cb->($handle, $fatal, $message) 154=item on_error => $cb->($handle, $fatal, $message)
138 155
139This is the error callback, which is called when, well, some error 156This is the error callback, which is called when, well, some error
140occured, such as not being able to resolve the hostname, failure to 157occured, such as not being able to resolve the hostname, failure to
141connect or a read error. 158connect, or a read error.
142 159
143Some errors are fatal (which is indicated by C<$fatal> being true). On 160Some errors are fatal (which is indicated by C<$fatal> being true). On
144fatal errors the handle object will be destroyed (by a call to C<< -> 161fatal errors the handle object will be destroyed (by a call to C<< ->
145destroy >>) after invoking the error callback (which means you are free to 162destroy >>) after invoking the error callback (which means you are free to
146examine the handle object). Examples of fatal errors are an EOF condition 163examine the handle object). Examples of fatal errors are an EOF condition
147with active (but unsatisifable) read watchers (C<EPIPE>) or I/O errors. In 164with active (but unsatisfiable) read watchers (C<EPIPE>) or I/O errors. In
148cases where the other side can close the connection at their will it is 165cases where the other side can close the connection at will, it is
149often easiest to not report C<EPIPE> errors in this callback. 166often easiest to not report C<EPIPE> errors in this callback.
150 167
151AnyEvent::Handle tries to find an appropriate error code for you to check 168AnyEvent::Handle tries to find an appropriate error code for you to check
152against, but in some cases (TLS errors), this does not work well. It is 169against, but in some cases (TLS errors), this does not work well. It is
153recommended to always output the C<$message> argument in human-readable 170recommended to always output the C<$message> argument in human-readable
154error messages (it's usually the same as C<"$!">). 171error messages (it's usually the same as C<"$!">).
155 172
156Non-fatal errors can be retried by simply returning, but it is recommended 173Non-fatal errors can be retried by returning, but it is recommended
157to simply ignore this parameter and instead abondon the handle object 174to simply ignore this parameter and instead abondon the handle object
158when this callback is invoked. Examples of non-fatal errors are timeouts 175when this callback is invoked. Examples of non-fatal errors are timeouts
159C<ETIMEDOUT>) or badly-formatted data (C<EBADMSG>). 176C<ETIMEDOUT>) or badly-formatted data (C<EBADMSG>).
160 177
161On callback entrance, the value of C<$!> contains the operating system 178On entry to the callback, the value of C<$!> contains the operating
162error code (or C<ENOSPC>, C<EPIPE>, C<ETIMEDOUT>, C<EBADMSG> or 179system error code (or C<ENOSPC>, C<EPIPE>, C<ETIMEDOUT>, C<EBADMSG> or
163C<EPROTO>). 180C<EPROTO>).
164 181
165While 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
166you will not be notified of errors otherwise. The default simply calls 183you will not be notified of errors otherwise. The default just calls
167C<croak>. 184C<croak>.
168 185
169=item on_read => $cb->($handle) 186=item on_read => $cb->($handle)
170 187
171This sets the default read callback, which is called when data arrives 188This sets the default read callback, which is called when data arrives
… …
176To access (and remove data from) the read buffer, use the C<< ->rbuf >> 193To access (and remove data from) the read buffer, use the C<< ->rbuf >>
177method or access the C<< $handle->{rbuf} >> member directly. Note that you 194method or access the C<< $handle->{rbuf} >> member directly. Note that you
178must not enlarge or modify the read buffer, you can only remove data at 195must not enlarge or modify the read buffer, you can only remove data at
179the beginning from it. 196the beginning from it.
180 197
198You can also call C<< ->push_read (...) >> or any other function that
199modifies the read queue. Or do both. Or ...
200
181When an EOF condition is detected then AnyEvent::Handle will first try to 201When an EOF condition is detected, AnyEvent::Handle will first try to
182feed all the remaining data to the queued callbacks and C<on_read> before 202feed all the remaining data to the queued callbacks and C<on_read> before
183calling the C<on_eof> callback. If no progress can be made, then a fatal 203calling the C<on_eof> callback. If no progress can be made, then a fatal
184error will be raised (with C<$!> set to C<EPIPE>). 204error will be raised (with C<$!> set to C<EPIPE>).
185 205
186Note that, unlike requests in the read queue, an C<on_read> callback 206Note that, unlike requests in the read queue, an C<on_read> callback
… …
205set, then a fatal error will be raised with C<$!> set to <0>. 225set, then a fatal error will be raised with C<$!> set to <0>.
206 226
207=item on_drain => $cb->($handle) 227=item on_drain => $cb->($handle)
208 228
209This 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
210(or when the callback is set and the buffer is empty already). 230(or immediately if the buffer is empty already).
211 231
212To append to the write buffer, use the C<< ->push_write >> method. 232To append to the write buffer, use the C<< ->push_write >> method.
213 233
214This 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
215into 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
… …
227many seconds pass without a successful read or write on the underlying 247many seconds pass without a successful read or write on the underlying
228file 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
229will 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>
230error will be raised). 250error will be raised).
231 251
232There are three variants of the timeouts that work fully independent 252There are three variants of the timeouts that work independently of each
233of each other, for both read and write, just read, and just write: 253other, for both read and write (triggered when nothing was read I<OR>
254written), just read (triggered when nothing was read), and just write:
234C<timeout>, C<rtimeout> and C<wtimeout>, with corresponding callbacks 255C<timeout>, C<rtimeout> and C<wtimeout>, with corresponding callbacks
235C<on_timeout>, C<on_rtimeout> and C<on_wtimeout>, and reset functions 256C<on_timeout>, C<on_rtimeout> and C<on_wtimeout>, and reset functions
236C<timeout_reset>, C<rtimeout_reset>, and C<wtimeout_reset>. 257C<timeout_reset>, C<rtimeout_reset>, and C<wtimeout_reset>.
237 258
238Note that timeout processing is also active when you currently do not have 259Note that timeout processing is active even when you do not have any
239any outstanding read or write requests: If you plan to keep the connection 260outstanding read or write requests: If you plan to keep the connection
240idle then you should disable the timout temporarily or ignore the timeout 261idle then you should disable the timeout temporarily or ignore the
241in the C<on_timeout> callback, in which case AnyEvent::Handle will simply 262timeout in the corresponding C<on_timeout> callback, in which case
242restart the timeout. 263AnyEvent::Handle will simply restart the timeout.
243 264
244Zero (the default) disables this timeout. 265Zero (the default) disables the corresponding timeout.
245 266
246=item on_timeout => $cb->($handle) 267=item on_timeout => $cb->($handle)
268
269=item on_rtimeout => $cb->($handle)
270
271=item on_wtimeout => $cb->($handle)
247 272
248Called whenever the inactivity timeout passes. If you return from this 273Called whenever the inactivity timeout passes. If you return from this
249callback, then the timeout will be reset as if some activity had happened, 274callback, then the timeout will be reset as if some activity had happened,
250so this condition is not fatal in any way. 275so this condition is not fatal in any way.
251 276
… …
259be configured to accept only so-and-so much data that it cannot act on 284be configured to accept only so-and-so much data that it cannot act on
260(for example, when expecting a line, an attacker could send an unlimited 285(for example, when expecting a line, an attacker could send an unlimited
261amount of data without a callback ever being called as long as the line 286amount of data without a callback ever being called as long as the line
262isn't finished). 287isn't finished).
263 288
289=item wbuf_max => <bytes>
290
291If defined, then a fatal error will be raised (with C<$!> set to C<ENOSPC>)
292when the write buffer ever (strictly) exceeds this size. This is useful to
293avoid some forms of denial-of-service attacks.
294
295Although the units of this parameter is bytes, this is the I<raw> number
296of bytes not yet accepted by the kernel. This can make a difference when
297you e.g. use TLS, as TLS typically makes your write data larger (but it
298can also make it smaller due to compression).
299
300As an example of when this limit is useful, take a chat server that sends
301chat messages to a client. If the client does not read those in a timely
302manner then the send buffer in the server would grow unbounded.
303
264=item autocork => <boolean> 304=item autocork => <boolean>
265 305
266When disabled (the default), then C<push_write> will try to immediately 306When disabled (the default), C<push_write> will try to immediately
267write the data to the handle, if possible. This avoids having to register 307write the data to the handle if possible. This avoids having to register
268a write watcher and wait for the next event loop iteration, but can 308a write watcher and wait for the next event loop iteration, but can
269be inefficient if you write multiple small chunks (on the wire, this 309be inefficient if you write multiple small chunks (on the wire, this
270disadvantage is usually avoided by your kernel's nagle algorithm, see 310disadvantage is usually avoided by your kernel's nagle algorithm, see
271C<no_delay>, but this option can save costly syscalls). 311C<no_delay>, but this option can save costly syscalls).
272 312
273When enabled, then writes will always be queued till the next event loop 313When enabled, writes will always be queued till the next event loop
274iteration. This is efficient when you do many small writes per iteration, 314iteration. This is efficient when you do many small writes per iteration,
275but less efficient when you do a single write only per iteration (or when 315but less efficient when you do a single write only per iteration (or when
276the write buffer often is full). It also increases write latency. 316the write buffer often is full). It also increases write latency.
277 317
278=item no_delay => <boolean> 318=item no_delay => <boolean>
… …
282the Nagle algorithm, and usually it is beneficial. 322the Nagle algorithm, and usually it is beneficial.
283 323
284In some situations you want as low a delay as possible, which can be 324In some situations you want as low a delay as possible, which can be
285accomplishd by setting this option to a true value. 325accomplishd by setting this option to a true value.
286 326
287The default is your opertaing system's default behaviour (most likely 327The default is your operating system's default behaviour (most likely
288enabled), this option explicitly enables or disables it, if possible. 328enabled). This option explicitly enables or disables it, if possible.
329
330=item keepalive => <boolean>
331
332Enables (default disable) the SO_KEEPALIVE option on the stream socket:
333normally, TCP connections have no time-out once established, so TCP
334connections, once established, can stay alive forever even when the other
335side has long gone. TCP keepalives are a cheap way to take down long-lived
336TCP connections when the other side becomes unreachable. While the default
337is OS-dependent, TCP keepalives usually kick in after around two hours,
338and, if the other side doesn't reply, take down the TCP connection some 10
339to 15 minutes later.
340
341It is harmless to specify this option for file handles that do not support
342keepalives, and enabling it on connections that are potentially long-lived
343is usually a good idea.
344
345=item oobinline => <boolean>
346
347BSD majorly fucked up the implementation of TCP urgent data. The result
348is that almost no OS implements TCP according to the specs, and every OS
349implements it slightly differently.
350
351If you want to handle TCP urgent data, then setting this flag (the default
352is enabled) gives you the most portable way of getting urgent data, by
353putting it into the stream.
354
355Since BSD emulation of OOB data on top of TCP's urgent data can have
356security implications, AnyEvent::Handle sets this flag automatically
357unless explicitly specified. Note that setting this flag after
358establishing a connection I<may> be a bit too late (data loss could
359already have occured on BSD systems), but at least it will protect you
360from most attacks.
289 361
290=item read_size => <bytes> 362=item read_size => <bytes>
291 363
292The default read block size (the amount of bytes this module will 364The initial read block size, the number of bytes this module will try to
293try to read during each loop iteration, which affects memory 365read during each loop iteration. Each handle object will consume at least
294requirements). Default: C<8192>. 366this amount of memory for the read buffer as well, so when handling many
367connections requirements). See also C<max_read_size>. Default: C<2048>.
368
369=item max_read_size => <bytes>
370
371The maximum read buffer size used by the dynamic adjustment
372algorithm: Each time AnyEvent::Handle can read C<read_size> bytes in
373one go it will double C<read_size> up to the maximum given by this
374option. Default: C<131072> or C<read_size>, whichever is higher.
295 375
296=item low_water_mark => <bytes> 376=item low_water_mark => <bytes>
297 377
298Sets the amount of bytes (default: C<0>) that make up an "empty" write 378Sets the number of bytes (default: C<0>) that make up an "empty" write
299buffer: If the write reaches this size or gets even samller it is 379buffer: If the buffer reaches this size or gets even samller it is
300considered empty. 380considered empty.
301 381
302Sometimes it can be beneficial (for performance reasons) to add data to 382Sometimes it can be beneficial (for performance reasons) to add data to
303the write buffer before it is fully drained, but this is a rare case, as 383the write buffer before it is fully drained, but this is a rare case, as
304the operating system kernel usually buffers data as well, so the default 384the operating system kernel usually buffers data as well, so the default
305is good in almost all cases. 385is good in almost all cases.
306 386
307=item linger => <seconds> 387=item linger => <seconds>
308 388
309If non-zero (default: C<3600>), then the destructor of the 389If this is non-zero (default: C<3600>), the destructor of the
310AnyEvent::Handle object will check whether there is still outstanding 390AnyEvent::Handle object will check whether there is still outstanding
311write data and will install a watcher that will write this data to the 391write data and will install a watcher that will write this data to the
312socket. No errors will be reported (this mostly matches how the operating 392socket. No errors will be reported (this mostly matches how the operating
313system treats outstanding data at socket close time). 393system treats outstanding data at socket close time).
314 394
… …
321A string used to identify the remote site - usually the DNS hostname 401A string used to identify the remote site - usually the DNS hostname
322(I<not> IDN!) used to create the connection, rarely the IP address. 402(I<not> IDN!) used to create the connection, rarely the IP address.
323 403
324Apart from being useful in error messages, this string is also used in TLS 404Apart from being useful in error messages, this string is also used in TLS
325peername verification (see C<verify_peername> in L<AnyEvent::TLS>). This 405peername verification (see C<verify_peername> in L<AnyEvent::TLS>). This
326verification will be skipped when C<peername> is not specified or 406verification will be skipped when C<peername> is not specified or is
327C<undef>. 407C<undef>.
328 408
329=item tls => "accept" | "connect" | Net::SSLeay::SSL object 409=item tls => "accept" | "connect" | Net::SSLeay::SSL object
330 410
331When this parameter is given, it enables TLS (SSL) mode, that means 411When this parameter is given, it enables TLS (SSL) mode, that means
332AnyEvent will start a TLS handshake as soon as the conenction has been 412AnyEvent will start a TLS handshake as soon as the connection has been
333established and will transparently encrypt/decrypt data afterwards. 413established and will transparently encrypt/decrypt data afterwards.
334 414
335All TLS protocol errors will be signalled as C<EPROTO>, with an 415All TLS protocol errors will be signalled as C<EPROTO>, with an
336appropriate error message. 416appropriate error message.
337 417
… …
357B<IMPORTANT:> since Net::SSLeay "objects" are really only integers, 437B<IMPORTANT:> since Net::SSLeay "objects" are really only integers,
358passing in the wrong integer will lead to certain crash. This most often 438passing in the wrong integer will lead to certain crash. This most often
359happens when one uses a stylish C<< tls => 1 >> and is surprised about the 439happens when one uses a stylish C<< tls => 1 >> and is surprised about the
360segmentation fault. 440segmentation fault.
361 441
362See the C<< ->starttls >> method for when need to start TLS negotiation later. 442Use the C<< ->starttls >> method if you need to start TLS negotiation later.
363 443
364=item tls_ctx => $anyevent_tls 444=item tls_ctx => $anyevent_tls
365 445
366Use the given C<AnyEvent::TLS> object to create the new TLS connection 446Use the given C<AnyEvent::TLS> object to create the new TLS connection
367(unless a connection object was specified directly). If this parameter is 447(unless a connection object was specified directly). If this
368missing, then AnyEvent::Handle will use C<AnyEvent::Handle::TLS_CTX>. 448parameter is missing (or C<undef>), then AnyEvent::Handle will use
449C<AnyEvent::Handle::TLS_CTX>.
369 450
370Instead of an object, you can also specify a hash reference with C<< key 451Instead of an object, you can also specify a hash reference with C<< key
371=> value >> pairs. Those will be passed to L<AnyEvent::TLS> to create a 452=> value >> pairs. Those will be passed to L<AnyEvent::TLS> to create a
372new TLS context object. 453new TLS context object.
373 454
… …
382 463
383TLS handshake failures will not cause C<on_error> to be invoked when this 464TLS handshake failures will not cause C<on_error> to be invoked when this
384callback is in effect, instead, the error message will be passed to C<on_starttls>. 465callback is in effect, instead, the error message will be passed to C<on_starttls>.
385 466
386Without this callback, handshake failures lead to C<on_error> being 467Without this callback, handshake failures lead to C<on_error> being
387called, as normal. 468called as usual.
388 469
389Note that you cannot call C<starttls> right again in this callback. If you 470Note that you cannot just call C<starttls> again in this callback. If you
390need to do that, start an zero-second timer instead whose callback can 471need to do that, start an zero-second timer instead whose callback can
391then call C<< ->starttls >> again. 472then call C<< ->starttls >> again.
392 473
393=item on_stoptls => $cb->($handle) 474=item on_stoptls => $cb->($handle)
394 475
… …
442 $self->{connect}[0], 523 $self->{connect}[0],
443 $self->{connect}[1], 524 $self->{connect}[1],
444 sub { 525 sub {
445 my ($fh, $host, $port, $retry) = @_; 526 my ($fh, $host, $port, $retry) = @_;
446 527
528 delete $self->{_connect}; # no longer needed
529
447 if ($fh) { 530 if ($fh) {
448 $self->{fh} = $fh; 531 $self->{fh} = $fh;
449 532
450 delete $self->{_skip_drain_rbuf}; 533 delete $self->{_skip_drain_rbuf};
451 $self->_start; 534 $self->_start;
452 535
453 $self->{on_connect} 536 $self->{on_connect}
454 and $self->{on_connect}($self, $host, $port, sub { 537 and $self->{on_connect}($self, $host, $port, sub {
455 delete @$self{qw(fh _tw _ww _rw _eof _queue rbuf _wbuf tls _tls_rbuf _tls_wbuf)}; 538 delete @$self{qw(fh _tw _rtw _wtw _ww _rw _eof _queue rbuf _wbuf tls _tls_rbuf _tls_wbuf)};
456 $self->{_skip_drain_rbuf} = 1; 539 $self->{_skip_drain_rbuf} = 1;
457 &$retry; 540 &$retry;
458 }); 541 });
459 542
460 } else { 543 } else {
461 if ($self->{on_connect_error}) { 544 if ($self->{on_connect_error}) {
462 $self->{on_connect_error}($self, "$!"); 545 $self->{on_connect_error}($self, "$!");
463 $self->destroy; 546 $self->destroy if $self;
464 } else { 547 } else {
465 $self->_error ($!, 1); 548 $self->_error ($!, 1);
466 } 549 }
467 } 550 }
468 }, 551 },
469 sub { 552 sub {
470 local $self->{fh} = $_[0]; 553 local $self->{fh} = $_[0];
471 554
472 $self->{on_prepare} 555 $self->{on_prepare}
473 ? $self->{on_prepare}->($self) 556 ? $self->{on_prepare}->($self)
474 : () 557 : ()
475 } 558 }
476 ); 559 );
477 } 560 }
478 561
… …
484} 567}
485 568
486sub _start { 569sub _start {
487 my ($self) = @_; 570 my ($self) = @_;
488 571
572 # too many clueless people try to use udp and similar sockets
573 # with AnyEvent::Handle, do them a favour.
574 my $type = getsockopt $self->{fh}, Socket::SOL_SOCKET (), Socket::SO_TYPE ();
575 Carp::croak "AnyEvent::Handle: only stream sockets supported, anything else will NOT work!"
576 if Socket::SOCK_STREAM () != (unpack "I", $type) && defined $type;
577
489 AnyEvent::Util::fh_nonblocking $self->{fh}, 1; 578 AnyEvent::Util::fh_nonblocking $self->{fh}, 1;
490 579
491 $self->{_activity} = 580 $self->{_activity} =
492 $self->{_ractivity} = 581 $self->{_ractivity} =
493 $self->{_wactivity} = AE::now; 582 $self->{_wactivity} = AE::now;
494 583
584 $self->{read_size} ||= 2048;
585 $self->{max_read_size} = $self->{read_size}
586 if $self->{read_size} > ($self->{max_read_size} || MAX_READ_SIZE);
587
495 $self->timeout (delete $self->{timeout} ) if $self->{timeout}; 588 $self->timeout (delete $self->{timeout} ) if $self->{timeout};
496 $self->rtimeout (delete $self->{rtimeout}) if $self->{rtimeout}; 589 $self->rtimeout (delete $self->{rtimeout} ) if $self->{rtimeout};
497 $self->wtimeout (delete $self->{wtimeout}) if $self->{wtimeout}; 590 $self->wtimeout (delete $self->{wtimeout} ) if $self->{wtimeout};
498 591
499 $self->no_delay (delete $self->{no_delay}) if exists $self->{no_delay}; 592 $self->no_delay (delete $self->{no_delay} ) if exists $self->{no_delay} && $self->{no_delay};
593 $self->keepalive (delete $self->{keepalive}) if exists $self->{keepalive} && $self->{keepalive};
500 594
595 $self->oobinline (exists $self->{oobinline} ? delete $self->{oobinline} : 1);
596
501 $self->starttls (delete $self->{tls}, delete $self->{tls_ctx}) 597 $self->starttls (delete $self->{tls}, delete $self->{tls_ctx})
502 if $self->{tls}; 598 if $self->{tls};
503 599
504 $self->on_drain (delete $self->{on_drain}) if $self->{on_drain}; 600 $self->on_drain (delete $self->{on_drain} ) if $self->{on_drain};
505 601
506 $self->start_read 602 $self->start_read
507 if $self->{on_read} || @{ $self->{_queue} }; 603 if $self->{on_read} || @{ $self->{_queue} };
508 604
509 $self->_drain_wbuf; 605 $self->_drain_wbuf;
510} 606}
511
512#sub _shutdown {
513# my ($self) = @_;
514#
515# delete @$self{qw(_tw _rw _ww fh wbuf on_read _queue)};
516# $self->{_eof} = 1; # tell starttls et. al to stop trying
517#
518# &_freetls;
519#}
520 607
521sub _error { 608sub _error {
522 my ($self, $errno, $fatal, $message) = @_; 609 my ($self, $errno, $fatal, $message) = @_;
523 610
524 $! = $errno; 611 $! = $errno;
525 $message ||= "$!"; 612 $message ||= "$!";
526 613
527 if ($self->{on_error}) { 614 if ($self->{on_error}) {
528 $self->{on_error}($self, $fatal, $message); 615 $self->{on_error}($self, $fatal, $message);
529 $self->destroy if $fatal; 616 $self->destroy if $fatal;
530 } elsif ($self->{fh}) { 617 } elsif ($self->{fh} || $self->{connect}) {
531 $self->destroy; 618 $self->destroy;
532 Carp::croak "AnyEvent::Handle uncaught error: $message"; 619 Carp::croak "AnyEvent::Handle uncaught error: $message";
533 } 620 }
534} 621}
535 622
… …
594=cut 681=cut
595 682
596sub no_delay { 683sub no_delay {
597 $_[0]{no_delay} = $_[1]; 684 $_[0]{no_delay} = $_[1];
598 685
686 setsockopt $_[0]{fh}, Socket::IPPROTO_TCP (), Socket::TCP_NODELAY (), int $_[1]
687 if $_[0]{fh};
688}
689
690=item $handle->keepalive ($boolean)
691
692Enables or disables the C<keepalive> setting (see constructor argument of
693the same name for details).
694
695=cut
696
697sub keepalive {
698 $_[0]{keepalive} = $_[1];
699
599 eval { 700 eval {
600 local $SIG{__DIE__}; 701 local $SIG{__DIE__};
601 setsockopt $_[0]{fh}, &Socket::IPPROTO_TCP, &Socket::TCP_NODELAY, int $_[1] 702 setsockopt $_[0]{fh}, Socket::SOL_SOCKET (), Socket::SO_KEEPALIVE (), int $_[1]
602 if $_[0]{fh}; 703 if $_[0]{fh};
603 }; 704 };
604} 705}
605 706
707=item $handle->oobinline ($boolean)
708
709Enables or disables the C<oobinline> setting (see constructor argument of
710the same name for details).
711
712=cut
713
714sub oobinline {
715 $_[0]{oobinline} = $_[1];
716
717 eval {
718 local $SIG{__DIE__};
719 setsockopt $_[0]{fh}, Socket::SOL_SOCKET (), Socket::SO_OOBINLINE (), int $_[1]
720 if $_[0]{fh};
721 };
722}
723
724=item $handle->keepalive ($boolean)
725
726Enables or disables the C<keepalive> setting (see constructor argument of
727the same name for details).
728
729=cut
730
731sub keepalive {
732 $_[0]{keepalive} = $_[1];
733
734 eval {
735 local $SIG{__DIE__};
736 setsockopt $_[0]{fh}, Socket::SOL_SOCKET (), Socket::SO_KEEPALIVE (), int $_[1]
737 if $_[0]{fh};
738 };
739}
740
606=item $handle->on_starttls ($cb) 741=item $handle->on_starttls ($cb)
607 742
608Replace the current C<on_starttls> callback (see the C<on_starttls> constructor argument). 743Replace the current C<on_starttls> callback (see the C<on_starttls> constructor argument).
609 744
610=cut 745=cut
… …
617 752
618Replace the current C<on_stoptls> callback (see the C<on_stoptls> constructor argument). 753Replace the current C<on_stoptls> callback (see the C<on_stoptls> constructor argument).
619 754
620=cut 755=cut
621 756
622sub on_starttls { 757sub on_stoptls {
623 $_[0]{on_stoptls} = $_[1]; 758 $_[0]{on_stoptls} = $_[1];
624} 759}
625 760
626=item $handle->rbuf_max ($max_octets) 761=item $handle->rbuf_max ($max_octets)
627 762
628Configures the C<rbuf_max> setting (C<undef> disables it). 763Configures the C<rbuf_max> setting (C<undef> disables it).
764
765=item $handle->wbuf_max ($max_octets)
766
767Configures the C<wbuf_max> setting (C<undef> disables it).
629 768
630=cut 769=cut
631 770
632sub rbuf_max { 771sub rbuf_max {
633 $_[0]{rbuf_max} = $_[1]; 772 $_[0]{rbuf_max} = $_[1];
634} 773}
635 774
775sub wbuf_max {
776 $_[0]{wbuf_max} = $_[1];
777}
778
636############################################################################# 779#############################################################################
637 780
638=item $handle->timeout ($seconds) 781=item $handle->timeout ($seconds)
639 782
640=item $handle->rtimeout ($seconds) 783=item $handle->rtimeout ($seconds)
641 784
642=item $handle->wtimeout ($seconds) 785=item $handle->wtimeout ($seconds)
643 786
644Configures (or disables) the inactivity timeout. 787Configures (or disables) the inactivity timeout.
788
789The timeout will be checked instantly, so this method might destroy the
790handle before it returns.
645 791
646=item $handle->timeout_reset 792=item $handle->timeout_reset
647 793
648=item $handle->rtimeout_reset 794=item $handle->rtimeout_reset
649 795
… …
666 $_[0]{$on_timeout} = $_[1]; 812 $_[0]{$on_timeout} = $_[1];
667 }; 813 };
668 814
669 *$timeout = sub { 815 *$timeout = sub {
670 my ($self, $new_value) = @_; 816 my ($self, $new_value) = @_;
817
818 $new_value >= 0
819 or Carp::croak "AnyEvent::Handle->$timeout called with negative timeout ($new_value), caught";
671 820
672 $self->{$timeout} = $new_value; 821 $self->{$timeout} = $new_value;
673 delete $self->{$tw}; &$cb; 822 delete $self->{$tw}; &$cb;
674 }; 823 };
675 824
… …
739=item $handle->on_drain ($cb) 888=item $handle->on_drain ($cb)
740 889
741Sets the C<on_drain> callback or clears it (see the description of 890Sets the C<on_drain> callback or clears it (see the description of
742C<on_drain> in the constructor). 891C<on_drain> in the constructor).
743 892
893This method may invoke callbacks (and therefore the handle might be
894destroyed after it returns).
895
744=cut 896=cut
745 897
746sub on_drain { 898sub on_drain {
747 my ($self, $cb) = @_; 899 my ($self, $cb) = @_;
748 900
… …
752 if $cb && $self->{low_water_mark} >= (length $self->{wbuf}) + (length $self->{_tls_wbuf}); 904 if $cb && $self->{low_water_mark} >= (length $self->{wbuf}) + (length $self->{_tls_wbuf});
753} 905}
754 906
755=item $handle->push_write ($data) 907=item $handle->push_write ($data)
756 908
757Queues the given scalar to be written. You can push as much data as you 909Queues the given scalar to be written. You can push as much data as
758want (only limited by the available memory), as C<AnyEvent::Handle> 910you want (only limited by the available memory and C<wbuf_max>), as
759buffers it independently of the kernel. 911C<AnyEvent::Handle> buffers it independently of the kernel.
912
913This method may invoke callbacks (and therefore the handle might be
914destroyed after it returns).
760 915
761=cut 916=cut
762 917
763sub _drain_wbuf { 918sub _drain_wbuf {
764 my ($self) = @_; 919 my ($self) = @_;
… …
789 $cb->() unless $self->{autocork}; 944 $cb->() unless $self->{autocork};
790 945
791 # if still data left in wbuf, we need to poll 946 # if still data left in wbuf, we need to poll
792 $self->{_ww} = AE::io $self->{fh}, 1, $cb 947 $self->{_ww} = AE::io $self->{fh}, 1, $cb
793 if length $self->{wbuf}; 948 if length $self->{wbuf};
949
950 if (
951 defined $self->{wbuf_max}
952 && $self->{wbuf_max} < length $self->{wbuf}
953 ) {
954 $self->_error (Errno::ENOSPC, 1), return;
955 }
794 }; 956 };
795} 957}
796 958
797our %WH; 959our %WH;
798 960
961# deprecated
799sub register_write_type($$) { 962sub register_write_type($$) {
800 $WH{$_[0]} = $_[1]; 963 $WH{$_[0]} = $_[1];
801} 964}
802 965
803sub push_write { 966sub push_write {
804 my $self = shift; 967 my $self = shift;
805 968
806 if (@_ > 1) { 969 if (@_ > 1) {
807 my $type = shift; 970 my $type = shift;
808 971
972 @_ = ($WH{$type} ||= _load_func "$type\::anyevent_write_type"
809 @_ = ($WH{$type} or Carp::croak "unsupported type passed to AnyEvent::Handle::push_write") 973 or Carp::croak "unsupported/unloadable type '$type' passed to AnyEvent::Handle::push_write")
810 ->($self, @_); 974 ->($self, @_);
811 } 975 }
812 976
977 # we downgrade here to avoid hard-to-track-down bugs,
978 # and diagnose the problem earlier and better.
979
813 if ($self->{tls}) { 980 if ($self->{tls}) {
814 $self->{_tls_wbuf} .= $_[0]; 981 utf8::downgrade $self->{_tls_wbuf} .= $_[0];
815 &_dotls ($self) if $self->{fh}; 982 &_dotls ($self) if $self->{fh};
816 } else { 983 } else {
817 $self->{wbuf} .= $_[0]; 984 utf8::downgrade $self->{wbuf} .= $_[0];
818 $self->_drain_wbuf if $self->{fh}; 985 $self->_drain_wbuf if $self->{fh};
819 } 986 }
820} 987}
821 988
822=item $handle->push_write (type => @args) 989=item $handle->push_write (type => @args)
823 990
824Instead of formatting your data yourself, you can also let this module do 991Instead of formatting your data yourself, you can also let this module
825the job by specifying a type and type-specific arguments. 992do the job by specifying a type and type-specific arguments. You
993can also specify the (fully qualified) name of a package, in which
994case AnyEvent tries to load the package and then expects to find the
995C<anyevent_write_type> function inside (see "custom write types", below).
826 996
827Predefined types are (if you have ideas for additional types, feel free to 997Predefined types are (if you have ideas for additional types, feel free to
828drop by and tell us): 998drop by and tell us):
829 999
830=over 4 1000=over 4
… …
887Other languages could read single lines terminated by a newline and pass 1057Other languages could read single lines terminated by a newline and pass
888this line into their JSON decoder of choice. 1058this line into their JSON decoder of choice.
889 1059
890=cut 1060=cut
891 1061
1062sub json_coder() {
1063 eval { require JSON::XS; JSON::XS->new->utf8 }
1064 || do { require JSON; JSON->new->utf8 }
1065}
1066
892register_write_type json => sub { 1067register_write_type json => sub {
893 my ($self, $ref) = @_; 1068 my ($self, $ref) = @_;
894 1069
895 require JSON; 1070 my $json = $self->{json} ||= json_coder;
896 1071
897 $self->{json} ? $self->{json}->encode ($ref) 1072 $json->encode ($ref)
898 : JSON::encode_json ($ref)
899}; 1073};
900 1074
901=item storable => $reference 1075=item storable => $reference
902 1076
903Freezes the given reference using L<Storable> and writes it to the 1077Freezes the given reference using L<Storable> and writes it to the
… …
921before it was actually written. One way to do that is to replace your 1095before it was actually written. One way to do that is to replace your
922C<on_drain> handler by a callback that shuts down the socket (and set 1096C<on_drain> handler by a callback that shuts down the socket (and set
923C<low_water_mark> to C<0>). This method is a shorthand for just that, and 1097C<low_water_mark> to C<0>). This method is a shorthand for just that, and
924replaces the C<on_drain> callback with: 1098replaces the C<on_drain> callback with:
925 1099
926 sub { shutdown $_[0]{fh}, 1 } # for push_shutdown 1100 sub { shutdown $_[0]{fh}, 1 }
927 1101
928This simply shuts down the write side and signals an EOF condition to the 1102This simply shuts down the write side and signals an EOF condition to the
929the peer. 1103the peer.
930 1104
931You can rely on the normal read queue and C<on_eof> handling 1105You can rely on the normal read queue and C<on_eof> handling
932afterwards. This is the cleanest way to close a connection. 1106afterwards. This is the cleanest way to close a connection.
933 1107
1108This method may invoke callbacks (and therefore the handle might be
1109destroyed after it returns).
1110
934=cut 1111=cut
935 1112
936sub push_shutdown { 1113sub push_shutdown {
937 my ($self) = @_; 1114 my ($self) = @_;
938 1115
939 delete $self->{low_water_mark}; 1116 delete $self->{low_water_mark};
940 $self->on_drain (sub { shutdown $_[0]{fh}, 1 }); 1117 $self->on_drain (sub { shutdown $_[0]{fh}, 1 });
941} 1118}
942 1119
943=item AnyEvent::Handle::register_write_type type => $coderef->($handle, @args) 1120=item custom write types - Package::anyevent_write_type $handle, @args
944 1121
945This function (not method) lets you add your own types to C<push_write>. 1122Instead of one of the predefined types, you can also specify the name of
1123a package. AnyEvent will try to load the package and then expects to find
1124a function named C<anyevent_write_type> inside. If it isn't found, it
1125progressively tries to load the parent package until it either finds the
1126function (good) or runs out of packages (bad).
1127
946Whenever the given C<type> is used, C<push_write> will invoke the code 1128Whenever the given C<type> is used, C<push_write> will the function with
947reference with the handle object and the remaining arguments. 1129the handle object and the remaining arguments.
948 1130
949The code reference is supposed to return a single octet string that will 1131The function is supposed to return a single octet string that will be
950be appended to the write buffer. 1132appended to the write buffer, so you cna mentally treat this function as a
1133"arguments to on-the-wire-format" converter.
951 1134
952Note that this is a function, and all types registered this way will be 1135Example: implement a custom write type C<join> that joins the remaining
953global, so try to use unique names. 1136arguments using the first one.
1137
1138 $handle->push_write (My::Type => " ", 1,2,3);
1139
1140 # uses the following package, which can be defined in the "My::Type" or in
1141 # the "My" modules to be auto-loaded, or just about anywhere when the
1142 # My::Type::anyevent_write_type is defined before invoking it.
1143
1144 package My::Type;
1145
1146 sub anyevent_write_type {
1147 my ($handle, $delim, @args) = @_;
1148
1149 join $delim, @args
1150 }
954 1151
955=cut 1152=cut
956 1153
957############################################################################# 1154#############################################################################
958 1155
… …
967ways, the "simple" way, using only C<on_read> and the "complex" way, using 1164ways, the "simple" way, using only C<on_read> and the "complex" way, using
968a queue. 1165a queue.
969 1166
970In the simple case, you just install an C<on_read> callback and whenever 1167In the simple case, you just install an C<on_read> callback and whenever
971new data arrives, it will be called. You can then remove some data (if 1168new data arrives, it will be called. You can then remove some data (if
972enough is there) from the read buffer (C<< $handle->rbuf >>). Or you cna 1169enough is there) from the read buffer (C<< $handle->rbuf >>). Or you can
973leave the data there if you want to accumulate more (e.g. when only a 1170leave the data there if you want to accumulate more (e.g. when only a
974partial message has been received so far). 1171partial message has been received so far), or change the read queue with
1172e.g. C<push_read>.
975 1173
976In the more complex case, you want to queue multiple callbacks. In this 1174In the more complex case, you want to queue multiple callbacks. In this
977case, AnyEvent::Handle will call the first queued callback each time new 1175case, AnyEvent::Handle will call the first queued callback each time new
978data arrives (also the first time it is queued) and removes it when it has 1176data arrives (also the first time it is queued) and remove it when it has
979done its job (see C<push_read>, below). 1177done its job (see C<push_read>, below).
980 1178
981This way you can, for example, push three line-reads, followed by reading 1179This way you can, for example, push three line-reads, followed by reading
982a chunk of data, and AnyEvent::Handle will execute them in order. 1180a chunk of data, and AnyEvent::Handle will execute them in order.
983 1181
… …
1114 1312
1115This replaces the currently set C<on_read> callback, or clears it (when 1313This replaces the currently set C<on_read> callback, or clears it (when
1116the new callback is C<undef>). See the description of C<on_read> in the 1314the new callback is C<undef>). See the description of C<on_read> in the
1117constructor. 1315constructor.
1118 1316
1317This method may invoke callbacks (and therefore the handle might be
1318destroyed after it returns).
1319
1119=cut 1320=cut
1120 1321
1121sub on_read { 1322sub on_read {
1122 my ($self, $cb) = @_; 1323 my ($self, $cb) = @_;
1123 1324
… …
1125 $self->_drain_rbuf if $cb; 1326 $self->_drain_rbuf if $cb;
1126} 1327}
1127 1328
1128=item $handle->rbuf 1329=item $handle->rbuf
1129 1330
1130Returns the read buffer (as a modifiable lvalue). 1331Returns the read buffer (as a modifiable lvalue). You can also access the
1332read buffer directly as the C<< ->{rbuf} >> member, if you want (this is
1333much faster, and no less clean).
1131 1334
1132You can access the read buffer directly as the C<< ->{rbuf} >> 1335The only operation allowed on the read buffer (apart from looking at it)
1133member, if you want. However, the only operation allowed on the 1336is removing data from its beginning. Otherwise modifying or appending to
1134read buffer (apart from looking at it) is removing data from its 1337it is not allowed and will lead to hard-to-track-down bugs.
1135beginning. Otherwise modifying or appending to it is not allowed and will
1136lead to hard-to-track-down bugs.
1137 1338
1138NOTE: The read buffer should only be used or modified if the C<on_read>, 1339NOTE: The read buffer should only be used or modified in the C<on_read>
1139C<push_read> or C<unshift_read> methods are used. The other read methods 1340callback or when C<push_read> or C<unshift_read> are used with a single
1140automatically manage the read buffer. 1341callback (i.e. untyped). Typed C<push_read> and C<unshift_read> methods
1342will manage the read buffer on their own.
1141 1343
1142=cut 1344=cut
1143 1345
1144sub rbuf : lvalue { 1346sub rbuf : lvalue {
1145 $_[0]{rbuf} 1347 $_[0]{rbuf}
… …
1162 1364
1163If enough data was available, then the callback must remove all data it is 1365If enough data was available, then the callback must remove all data it is
1164interested in (which can be none at all) and return a true value. After returning 1366interested in (which can be none at all) and return a true value. After returning
1165true, it will be removed from the queue. 1367true, it will be removed from the queue.
1166 1368
1369These methods may invoke callbacks (and therefore the handle might be
1370destroyed after it returns).
1371
1167=cut 1372=cut
1168 1373
1169our %RH; 1374our %RH;
1170 1375
1171sub register_read_type($$) { 1376sub register_read_type($$) {
… …
1177 my $cb = pop; 1382 my $cb = pop;
1178 1383
1179 if (@_) { 1384 if (@_) {
1180 my $type = shift; 1385 my $type = shift;
1181 1386
1387 $cb = ($RH{$type} ||= _load_func "$type\::anyevent_read_type"
1182 $cb = ($RH{$type} or Carp::croak "unsupported type passed to AnyEvent::Handle::push_read") 1388 or Carp::croak "unsupported/unloadable type '$type' passed to AnyEvent::Handle::push_read")
1183 ->($self, $cb, @_); 1389 ->($self, $cb, @_);
1184 } 1390 }
1185 1391
1186 push @{ $self->{_queue} }, $cb; 1392 push @{ $self->{_queue} }, $cb;
1187 $self->_drain_rbuf; 1393 $self->_drain_rbuf;
… …
1192 my $cb = pop; 1398 my $cb = pop;
1193 1399
1194 if (@_) { 1400 if (@_) {
1195 my $type = shift; 1401 my $type = shift;
1196 1402
1403 $cb = ($RH{$type} ||= _load_func "$type\::anyevent_read_type"
1197 $cb = ($RH{$type} or Carp::croak "unsupported type passed to AnyEvent::Handle::unshift_read") 1404 or Carp::croak "unsupported/unloadable type '$type' passed to AnyEvent::Handle::unshift_read")
1198 ->($self, $cb, @_); 1405 ->($self, $cb, @_);
1199 } 1406 }
1200
1201 1407
1202 unshift @{ $self->{_queue} }, $cb; 1408 unshift @{ $self->{_queue} }, $cb;
1203 $self->_drain_rbuf; 1409 $self->_drain_rbuf;
1204} 1410}
1205 1411
… …
1207 1413
1208=item $handle->unshift_read (type => @args, $cb) 1414=item $handle->unshift_read (type => @args, $cb)
1209 1415
1210Instead of providing a callback that parses the data itself you can chose 1416Instead of providing a callback that parses the data itself you can chose
1211between a number of predefined parsing formats, for chunks of data, lines 1417between a number of predefined parsing formats, for chunks of data, lines
1212etc. 1418etc. You can also specify the (fully qualified) name of a package, in
1419which case AnyEvent tries to load the package and then expects to find the
1420C<anyevent_read_type> function inside (see "custom read types", below).
1213 1421
1214Predefined types are (if you have ideas for additional types, feel free to 1422Predefined types are (if you have ideas for additional types, feel free to
1215drop by and tell us): 1423drop by and tell us):
1216 1424
1217=over 4 1425=over 4
… …
1309the receive buffer when neither C<$accept> nor C<$reject> match, 1517the receive buffer when neither C<$accept> nor C<$reject> match,
1310and everything preceding and including the match will be accepted 1518and everything preceding and including the match will be accepted
1311unconditionally. This is useful to skip large amounts of data that you 1519unconditionally. This is useful to skip large amounts of data that you
1312know cannot be matched, so that the C<$accept> or C<$reject> regex do not 1520know cannot be matched, so that the C<$accept> or C<$reject> regex do not
1313have to start matching from the beginning. This is purely an optimisation 1521have to start matching from the beginning. This is purely an optimisation
1314and is usually worth only when you expect more than a few kilobytes. 1522and is usually worth it only when you expect more than a few kilobytes.
1315 1523
1316Example: expect a http header, which ends at C<\015\012\015\012>. Since we 1524Example: expect a http header, which ends at C<\015\012\015\012>. Since we
1317expect the header to be very large (it isn't in practise, but...), we use 1525expect the header to be very large (it isn't in practice, but...), we use
1318a skip regex to skip initial portions. The skip regex is tricky in that 1526a skip regex to skip initial portions. The skip regex is tricky in that
1319it only accepts something not ending in either \015 or \012, as these are 1527it only accepts something not ending in either \015 or \012, as these are
1320required for the accept regex. 1528required for the accept regex.
1321 1529
1322 $handle->push_read (regex => 1530 $handle->push_read (regex =>
… …
1457=cut 1665=cut
1458 1666
1459register_read_type json => sub { 1667register_read_type json => sub {
1460 my ($self, $cb) = @_; 1668 my ($self, $cb) = @_;
1461 1669
1462 my $json = $self->{json} ||= 1670 my $json = $self->{json} ||= json_coder;
1463 eval { require JSON::XS; JSON::XS->new->utf8 }
1464 || do { require JSON; JSON->new->utf8 };
1465 1671
1466 my $data; 1672 my $data;
1467 my $rbuf = \$self->{rbuf}; 1673 my $rbuf = \$self->{rbuf};
1468 1674
1469 sub { 1675 sub {
… …
1538 } 1744 }
1539}; 1745};
1540 1746
1541=back 1747=back
1542 1748
1543=item AnyEvent::Handle::register_read_type type => $coderef->($handle, $cb, @args) 1749=item custom read types - Package::anyevent_read_type $handle, $cb, @args
1544 1750
1545This function (not method) lets you add your own types to C<push_read>. 1751Instead of one of the predefined types, you can also specify the name
1752of a package. AnyEvent will try to load the package and then expects to
1753find a function named C<anyevent_read_type> inside. If it isn't found, it
1754progressively tries to load the parent package until it either finds the
1755function (good) or runs out of packages (bad).
1546 1756
1547Whenever the given C<type> is used, C<push_read> will invoke the code 1757Whenever this type is used, C<push_read> will invoke the function with the
1548reference with the handle object, the callback and the remaining 1758handle object, the original callback and the remaining arguments.
1549arguments.
1550 1759
1551The code reference is supposed to return a callback (usually a closure) 1760The function is supposed to return a callback (usually a closure) that
1552that works as a plain read callback (see C<< ->push_read ($cb) >>). 1761works as a plain read callback (see C<< ->push_read ($cb) >>), so you can
1762mentally treat the function as a "configurable read type to read callback"
1763converter.
1553 1764
1554It should invoke the passed callback when it is done reading (remember to 1765It should invoke the original callback when it is done reading (remember
1555pass C<$handle> as first argument as all other callbacks do that). 1766to pass C<$handle> as first argument as all other callbacks do that,
1767although there is no strict requirement on this).
1556 1768
1557Note that this is a function, and all types registered this way will be
1558global, so try to use unique names.
1559
1560For examples, see the source of this module (F<perldoc -m AnyEvent::Handle>, 1769For examples, see the source of this module (F<perldoc -m
1561search for C<register_read_type>)). 1770AnyEvent::Handle>, search for C<register_read_type>)).
1562 1771
1563=item $handle->stop_read 1772=item $handle->stop_read
1564 1773
1565=item $handle->start_read 1774=item $handle->start_read
1566 1775
… …
1572Note that AnyEvent::Handle will automatically C<start_read> for you when 1781Note that AnyEvent::Handle will automatically C<start_read> for you when
1573you change the C<on_read> callback or push/unshift a read callback, and it 1782you change the C<on_read> callback or push/unshift a read callback, and it
1574will automatically C<stop_read> for you when neither C<on_read> is set nor 1783will automatically C<stop_read> for you when neither C<on_read> is set nor
1575there are any read requests in the queue. 1784there are any read requests in the queue.
1576 1785
1577These methods will have no effect when in TLS mode (as TLS doesn't support 1786In older versions of this module (<= 5.3), these methods had no effect,
1578half-duplex connections). 1787as TLS does not support half-duplex connections. In current versions they
1788work as expected, as this behaviour is required to avoid certain resource
1789attacks, where the program would be forced to read (and buffer) arbitrary
1790amounts of data before being able to send some data. The drawback is that
1791some readings of the the SSL/TLS specifications basically require this
1792attack to be working, as SSL/TLS implementations might stall sending data
1793during a rehandshake.
1794
1795As a guideline, during the initial handshake, you should not stop reading,
1796and as a client, it might cause problems, depending on your applciation.
1579 1797
1580=cut 1798=cut
1581 1799
1582sub stop_read { 1800sub stop_read {
1583 my ($self) = @_; 1801 my ($self) = @_;
1584 1802
1585 delete $self->{_rw} unless $self->{tls}; 1803 delete $self->{_rw};
1586} 1804}
1587 1805
1588sub start_read { 1806sub start_read {
1589 my ($self) = @_; 1807 my ($self) = @_;
1590 1808
1591 unless ($self->{_rw} || $self->{_eof}) { 1809 unless ($self->{_rw} || $self->{_eof} || !$self->{fh}) {
1592 Scalar::Util::weaken $self; 1810 Scalar::Util::weaken $self;
1593 1811
1594 $self->{_rw} = AE::io $self->{fh}, 0, sub { 1812 $self->{_rw} = AE::io $self->{fh}, 0, sub {
1595 my $rbuf = \($self->{tls} ? my $buf : $self->{rbuf}); 1813 my $rbuf = \($self->{tls} ? my $buf : $self->{rbuf});
1596 my $len = sysread $self->{fh}, $$rbuf, $self->{read_size} || 8192, length $$rbuf; 1814 my $len = sysread $self->{fh}, $$rbuf, $self->{read_size}, length $$rbuf;
1597 1815
1598 if ($len > 0) { 1816 if ($len > 0) {
1599 $self->{_activity} = $self->{_ractivity} = AE::now; 1817 $self->{_activity} = $self->{_ractivity} = AE::now;
1600 1818
1601 if ($self->{tls}) { 1819 if ($self->{tls}) {
1602 Net::SSLeay::BIO_write ($self->{_rbio}, $$rbuf); 1820 Net::SSLeay::BIO_write ($self->{_rbio}, $$rbuf);
1603 1821
1604 &_dotls ($self); 1822 &_dotls ($self);
1605 } else { 1823 } else {
1606 $self->_drain_rbuf; 1824 $self->_drain_rbuf;
1825 }
1826
1827 if ($len == $self->{read_size}) {
1828 $self->{read_size} *= 2;
1829 $self->{read_size} = $self->{max_read_size} || MAX_READ_SIZE
1830 if $self->{read_size} > ($self->{max_read_size} || MAX_READ_SIZE);
1607 } 1831 }
1608 1832
1609 } elsif (defined $len) { 1833 } elsif (defined $len) {
1610 delete $self->{_rw}; 1834 delete $self->{_rw};
1611 $self->{_eof} = 1; 1835 $self->{_eof} = 1;
… …
1689 && ($tmp != $ERROR_SYSCALL || $!); 1913 && ($tmp != $ERROR_SYSCALL || $!);
1690 1914
1691 while (length ($tmp = Net::SSLeay::BIO_read ($self->{_wbio}))) { 1915 while (length ($tmp = Net::SSLeay::BIO_read ($self->{_wbio}))) {
1692 $self->{wbuf} .= $tmp; 1916 $self->{wbuf} .= $tmp;
1693 $self->_drain_wbuf; 1917 $self->_drain_wbuf;
1918 $self->{tls} or return; # tls session might have gone away in callback
1694 } 1919 }
1695 1920
1696 $self->{_on_starttls} 1921 $self->{_on_starttls}
1697 and Net::SSLeay::state ($self->{tls}) == Net::SSLeay::ST_OK () 1922 and Net::SSLeay::state ($self->{tls}) == Net::SSLeay::ST_OK ()
1698 and (delete $self->{_on_starttls})->($self, 1, "TLS/SSL connection established"); 1923 and (delete $self->{_on_starttls})->($self, 1, "TLS/SSL connection established");
… …
1720context in C<< $handle->{tls_ctx} >> after this call and can be used or 1945context in C<< $handle->{tls_ctx} >> after this call and can be used or
1721changed to your liking. Note that the handshake might have already started 1946changed to your liking. Note that the handshake might have already started
1722when this function returns. 1947when this function returns.
1723 1948
1724Due to bugs in OpenSSL, it might or might not be possible to do multiple 1949Due to bugs in OpenSSL, it might or might not be possible to do multiple
1725handshakes on the same stream. Best do not attempt to use the stream after 1950handshakes on the same stream. It is best to not attempt to use the
1726stopping TLS. 1951stream after stopping TLS.
1952
1953This method may invoke callbacks (and therefore the handle might be
1954destroyed after it returns).
1727 1955
1728=cut 1956=cut
1729 1957
1730our %TLS_CACHE; #TODO not yet documented, should we? 1958our %TLS_CACHE; #TODO not yet documented, should we?
1731 1959
… …
1743 require Net::SSLeay; 1971 require Net::SSLeay;
1744 1972
1745 $ERROR_SYSCALL = Net::SSLeay::ERROR_SYSCALL (); 1973 $ERROR_SYSCALL = Net::SSLeay::ERROR_SYSCALL ();
1746 $ERROR_WANT_READ = Net::SSLeay::ERROR_WANT_READ (); 1974 $ERROR_WANT_READ = Net::SSLeay::ERROR_WANT_READ ();
1747 1975
1748 $tls = $self->{tls}; 1976 $tls = delete $self->{tls};
1749 $ctx = $self->{tls_ctx}; 1977 $ctx = $self->{tls_ctx};
1750 1978
1751 local $Carp::CarpLevel = 1; # skip ourselves when creating a new context or session 1979 local $Carp::CarpLevel = 1; # skip ourselves when creating a new context or session
1752 1980
1753 if ("HASH" eq ref $ctx) { 1981 if ("HASH" eq ref $ctx) {
… …
1797 2025
1798=item $handle->stoptls 2026=item $handle->stoptls
1799 2027
1800Shuts down the SSL connection - this makes a proper EOF handshake by 2028Shuts down the SSL connection - this makes a proper EOF handshake by
1801sending a close notify to the other side, but since OpenSSL doesn't 2029sending a close notify to the other side, but since OpenSSL doesn't
1802support non-blocking shut downs, it is not guarenteed that you can re-use 2030support non-blocking shut downs, it is not guaranteed that you can re-use
1803the stream afterwards. 2031the stream afterwards.
2032
2033This method may invoke callbacks (and therefore the handle might be
2034destroyed after it returns).
1804 2035
1805=cut 2036=cut
1806 2037
1807sub stoptls { 2038sub stoptls {
1808 my ($self) = @_; 2039 my ($self) = @_;
1809 2040
1810 if ($self->{tls}) { 2041 if ($self->{tls} && $self->{fh}) {
1811 Net::SSLeay::shutdown ($self->{tls}); 2042 Net::SSLeay::shutdown ($self->{tls});
1812 2043
1813 &_dotls; 2044 &_dotls;
1814 2045
1815# # we don't give a shit. no, we do, but we can't. no...#d# 2046# # we don't give a shit. no, we do, but we can't. no...#d#
… …
1827 if $self->{tls} > 0; 2058 if $self->{tls} > 0;
1828 2059
1829 delete @$self{qw(_rbio _wbio _tls_wbuf _on_starttls)}; 2060 delete @$self{qw(_rbio _wbio _tls_wbuf _on_starttls)};
1830} 2061}
1831 2062
2063=item $handle->resettls
2064
2065This rarely-used method simply resets and TLS state on the handle, usually
2066causing data loss.
2067
2068One case where it may be useful is when you want to skip over the data in
2069the stream but you are not interested in interpreting it, so data loss is
2070no concern.
2071
2072=cut
2073
2074*resettls = \&_freetls;
2075
1832sub DESTROY { 2076sub DESTROY {
1833 my ($self) = @_; 2077 my ($self) = @_;
1834 2078
1835 &_freetls; 2079 &_freetls;
1836 2080
… …
1845 push @linger, AE::io $fh, 1, sub { 2089 push @linger, AE::io $fh, 1, sub {
1846 my $len = syswrite $fh, $wbuf, length $wbuf; 2090 my $len = syswrite $fh, $wbuf, length $wbuf;
1847 2091
1848 if ($len > 0) { 2092 if ($len > 0) {
1849 substr $wbuf, 0, $len, ""; 2093 substr $wbuf, 0, $len, "";
1850 } else { 2094 } elsif (defined $len || ($! != EAGAIN && $! != EINTR && $! != WSAEWOULDBLOCK)) {
1851 @linger = (); # end 2095 @linger = (); # end
1852 } 2096 }
1853 }; 2097 };
1854 push @linger, AE::timer $linger, 0, sub { 2098 push @linger, AE::timer $linger, 0, sub {
1855 @linger = (); 2099 @linger = ();
… …
1892 2136
1893sub AnyEvent::Handle::destroyed::AUTOLOAD { 2137sub AnyEvent::Handle::destroyed::AUTOLOAD {
1894 #nop 2138 #nop
1895} 2139}
1896 2140
2141=item $handle->destroyed
2142
2143Returns false as long as the handle hasn't been destroyed by a call to C<<
2144->destroy >>, true otherwise.
2145
2146Can be useful to decide whether the handle is still valid after some
2147callback possibly destroyed the handle. For example, C<< ->push_write >>,
2148C<< ->starttls >> and other methods can call user callbacks, which in turn
2149can destroy the handle, so work can be avoided by checking sometimes:
2150
2151 $hdl->starttls ("accept");
2152 return if $hdl->destroyed;
2153 $hdl->push_write (...
2154
2155Note that the call to C<push_write> will silently be ignored if the handle
2156has been destroyed, so often you can just ignore the possibility of the
2157handle being destroyed.
2158
2159=cut
2160
2161sub destroyed { 0 }
2162sub AnyEvent::Handle::destroyed::destroyed { 1 }
2163
1897=item AnyEvent::Handle::TLS_CTX 2164=item AnyEvent::Handle::TLS_CTX
1898 2165
1899This function creates and returns the AnyEvent::TLS object used by default 2166This function creates and returns the AnyEvent::TLS object used by default
1900for TLS mode. 2167for TLS mode.
1901 2168
… …
1928 2195
1929It is only safe to "forget" the reference inside EOF or error callbacks, 2196It is only safe to "forget" the reference inside EOF or error callbacks,
1930from within all other callbacks, you need to explicitly call the C<< 2197from within all other callbacks, you need to explicitly call the C<<
1931->destroy >> method. 2198->destroy >> method.
1932 2199
2200=item Why is my C<on_eof> callback never called?
2201
2202Probably because your C<on_error> callback is being called instead: When
2203you have outstanding requests in your read queue, then an EOF is
2204considered an error as you clearly expected some data.
2205
2206To avoid this, make sure you have an empty read queue whenever your handle
2207is supposed to be "idle" (i.e. connection closes are O.K.). You cna set
2208an C<on_read> handler that simply pushes the first read requests in the
2209queue.
2210
2211See also the next question, which explains this in a bit more detail.
2212
2213=item How can I serve requests in a loop?
2214
2215Most protocols consist of some setup phase (authentication for example)
2216followed by a request handling phase, where the server waits for requests
2217and handles them, in a loop.
2218
2219There are two important variants: The first (traditional, better) variant
2220handles requests until the server gets some QUIT command, causing it to
2221close the connection first (highly desirable for a busy TCP server). A
2222client dropping the connection is an error, which means this variant can
2223detect an unexpected detection close.
2224
2225To handle this case, always make sure you have a on-empty read queue, by
2226pushing the "read request start" handler on it:
2227
2228 # we assume a request starts with a single line
2229 my @start_request; @start_request = (line => sub {
2230 my ($hdl, $line) = @_;
2231
2232 ... handle request
2233
2234 # push next request read, possibly from a nested callback
2235 $hdl->push_read (@start_request);
2236 });
2237
2238 # auth done, now go into request handling loop
2239 # now push the first @start_request
2240 $hdl->push_read (@start_request);
2241
2242By always having an outstanding C<push_read>, the handle always expects
2243some data and raises the C<EPIPE> error when the connction is dropped
2244unexpectedly.
2245
2246The second variant is a protocol where the client can drop the connection
2247at any time. For TCP, this means that the server machine may run out of
2248sockets easier, and in general, it means you cnanot distinguish a protocl
2249failure/client crash from a normal connection close. Nevertheless, these
2250kinds of protocols are common (and sometimes even the best solution to the
2251problem).
2252
2253Having an outstanding read request at all times is possible if you ignore
2254C<EPIPE> errors, but this doesn't help with when the client drops the
2255connection during a request, which would still be an error.
2256
2257A better solution is to push the initial request read in an C<on_read>
2258callback. This avoids an error, as when the server doesn't expect data
2259(i.e. is idly waiting for the next request, an EOF will not raise an
2260error, but simply result in an C<on_eof> callback. It is also a bit slower
2261and simpler:
2262
2263 # auth done, now go into request handling loop
2264 $hdl->on_read (sub {
2265 my ($hdl) = @_;
2266
2267 # called each time we receive data but the read queue is empty
2268 # simply start read the request
2269
2270 $hdl->push_read (line => sub {
2271 my ($hdl, $line) = @_;
2272
2273 ... handle request
2274
2275 # do nothing special when the request has been handled, just
2276 # let the request queue go empty.
2277 });
2278 });
2279
1933=item I get different callback invocations in TLS mode/Why can't I pause 2280=item I get different callback invocations in TLS mode/Why can't I pause
1934reading? 2281reading?
1935 2282
1936Unlike, say, TCP, TLS connections do not consist of two independent 2283Unlike, say, TCP, TLS connections do not consist of two independent
1937communication channels, one for each direction. Or put differently. The 2284communication channels, one for each direction. Or put differently, the
1938read and write directions are not independent of each other: you cannot 2285read and write directions are not independent of each other: you cannot
1939write data unless you are also prepared to read, and vice versa. 2286write data unless you are also prepared to read, and vice versa.
1940 2287
1941This can mean than, in TLS mode, you might get C<on_error> or C<on_eof> 2288This means that, in TLS mode, you might get C<on_error> or C<on_eof>
1942callback invocations when you are not expecting any read data - the reason 2289callback invocations when you are not expecting any read data - the reason
1943is that AnyEvent::Handle always reads in TLS mode. 2290is that AnyEvent::Handle always reads in TLS mode.
1944 2291
1945During the connection, you have to make sure that you always have a 2292During the connection, you have to make sure that you always have a
1946non-empty read-queue, or an C<on_read> watcher. At the end of the 2293non-empty read-queue, or an C<on_read> watcher. At the end of the
… …
1960 my $data = delete $_[0]{rbuf}; 2307 my $data = delete $_[0]{rbuf};
1961 }); 2308 });
1962 2309
1963The reason to use C<on_error> is that TCP connections, due to latencies 2310The reason to use C<on_error> is that TCP connections, due to latencies
1964and packets loss, might get closed quite violently with an error, when in 2311and packets loss, might get closed quite violently with an error, when in
1965fact, all data has been received. 2312fact all data has been received.
1966 2313
1967It is usually better to use acknowledgements when transferring data, 2314It is usually better to use acknowledgements when transferring data,
1968to make sure the other side hasn't just died and you got the data 2315to make sure the other side hasn't just died and you got the data
1969intact. This is also one reason why so many internet protocols have an 2316intact. This is also one reason why so many internet protocols have an
1970explicit QUIT command. 2317explicit QUIT command.
… …
1987consider using C<< ->push_shutdown >> instead. 2334consider using C<< ->push_shutdown >> instead.
1988 2335
1989=item I want to contact a TLS/SSL server, I don't care about security. 2336=item I want to contact a TLS/SSL server, I don't care about security.
1990 2337
1991If your TLS server is a pure TLS server (e.g. HTTPS) that only speaks TLS, 2338If your TLS server is a pure TLS server (e.g. HTTPS) that only speaks TLS,
1992simply connect to it and then create the AnyEvent::Handle with the C<tls> 2339connect to it and then create the AnyEvent::Handle with the C<tls>
1993parameter: 2340parameter:
1994 2341
1995 tcp_connect $host, $port, sub { 2342 tcp_connect $host, $port, sub {
1996 my ($fh) = @_; 2343 my ($fh) = @_;
1997 2344
… …
2097 2444
2098=item * all members not documented here and not prefixed with an underscore 2445=item * all members not documented here and not prefixed with an underscore
2099are free to use in subclasses. 2446are free to use in subclasses.
2100 2447
2101Of course, new versions of AnyEvent::Handle may introduce more "public" 2448Of course, new versions of AnyEvent::Handle may introduce more "public"
2102member variables, but thats just life, at least it is documented. 2449member variables, but that's just life. At least it is documented.
2103 2450
2104=back 2451=back
2105 2452
2106=head1 AUTHOR 2453=head1 AUTHOR
2107 2454

Diff Legend

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