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 1.24 by root, Sat May 24 15:11:46 2008 UTC vs.
Revision 1.182 by root, Thu Sep 3 12:35:01 2009 UTC

1package AnyEvent::Handle;
2
3no warnings;
4use strict;
5
6use AnyEvent ();
7use AnyEvent::Util ();
8use Scalar::Util ();
9use Carp ();
10use Fcntl ();
11use Errno qw/EAGAIN EINTR/;
12
13=head1 NAME 1=head1 NAME
14 2
15AnyEvent::Handle - non-blocking I/O on file handles via AnyEvent 3AnyEvent::Handle - non-blocking I/O on file handles via AnyEvent
16
17This module is experimental.
18
19=cut
20
21our $VERSION = '0.04';
22 4
23=head1 SYNOPSIS 5=head1 SYNOPSIS
24 6
25 use AnyEvent; 7 use AnyEvent;
26 use AnyEvent::Handle; 8 use AnyEvent::Handle;
27 9
28 my $cv = AnyEvent->condvar; 10 my $cv = AnyEvent->condvar;
29 11
30 my $ae_fh = AnyEvent::Handle->new (fh => \*STDIN); 12 my $hdl; $hdl = new AnyEvent::Handle
31
32 #TODO
33
34 # or use the constructor to pass the callback:
35
36 my $ae_fh2 =
37 AnyEvent::Handle->new (
38 fh => \*STDIN, 13 fh => \*STDIN,
39 on_eof => sub { 14 on_error => sub {
40 $cv->broadcast; 15 my ($hdl, $fatal, $msg) = @_;
41 }, 16 warn "got error $msg\n";
42 #TODO 17 $hdl->destroy;
18 $cv->send;
43 ); 19 );
44 20
45 $cv->wait; 21 # send some request line
22 $hdl->push_write ("getinfo\015\012");
23
24 # read the response line
25 $hdl->push_read (line => sub {
26 my ($hdl, $line) = @_;
27 warn "got line <$line>\n";
28 $cv->send;
29 });
30
31 $cv->recv;
46 32
47=head1 DESCRIPTION 33=head1 DESCRIPTION
48 34
49This module is a helper module to make it easier to do event-based I/O on 35This module is a helper module to make it easier to do event-based I/O on
50filehandles. For utility functions for doing non-blocking connects and accepts 36filehandles.
51on sockets see L<AnyEvent::Util>. 37
38The L<AnyEvent::Intro> tutorial contains some well-documented
39AnyEvent::Handle examples.
52 40
53In the following, when the documentation refers to of "bytes" then this 41In the following, when the documentation refers to of "bytes" then this
54means characters. As sysread and syswrite are used for all I/O, their 42means characters. As sysread and syswrite are used for all I/O, their
55treatment of characters applies to this module as well. 43treatment of characters applies to this module as well.
56 44
45At the very minimum, you should specify C<fh> or C<connect>, and the
46C<on_error> callback.
47
57All callbacks will be invoked with the handle object as their first 48All callbacks will be invoked with the handle object as their first
58argument. 49argument.
59 50
51=cut
52
53package AnyEvent::Handle;
54
55use Scalar::Util ();
56use List::Util ();
57use Carp ();
58use Errno qw(EAGAIN EINTR);
59
60use AnyEvent (); BEGIN { AnyEvent::common_sense }
61use AnyEvent::Util qw(WSAEWOULDBLOCK);
62
63our $VERSION = $AnyEvent::VERSION;
64
60=head1 METHODS 65=head1 METHODS
61 66
62=over 4 67=over 4
63 68
64=item B<new (%args)> 69=item $handle = B<new> AnyEvent::TLS fh => $filehandle, key => value...
65 70
66The constructor supports these arguments (all as key => value pairs). 71The constructor supports these arguments (all as C<< key => value >> pairs).
67 72
68=over 4 73=over 4
69 74
70=item fh => $filehandle [MANDATORY] 75=item fh => $filehandle [C<fh> or C<connect> MANDATORY]
71 76
72The filehandle this L<AnyEvent::Handle> object will operate on. 77The filehandle this L<AnyEvent::Handle> object will operate on.
73
74NOTE: The filehandle will be set to non-blocking (using 78NOTE: The filehandle will be set to non-blocking mode (using
75AnyEvent::Util::fh_nonblocking). 79C<AnyEvent::Util::fh_nonblocking>) by the constructor and needs to stay in
80that mode.
76 81
77=item on_eof => $cb->($self) 82=item connect => [$host, $service] [C<fh> or C<connect> MANDATORY]
78 83
79Set the callback to be called on EOF. 84Try to connect to the specified host and service (port), using
85C<AnyEvent::Socket::tcp_connect>. The C<$host> additionally becomes the
86default C<peername>.
80 87
81While not mandatory, it is highly recommended to set an eof callback, 88You have to specify either this parameter, or C<fh>, above.
82otherwise you might end up with a closed socket while you are still
83waiting for data.
84 89
90It is possible to push requests on the read and write queues, and modify
91properties of the stream, even while AnyEvent::Handle is connecting.
92
93When this parameter is specified, then the C<on_prepare>,
94C<on_connect_error> and C<on_connect> callbacks will be called under the
95appropriate circumstances:
96
97=over 4
98
85=item on_error => $cb->($self) 99=item on_prepare => $cb->($handle)
86 100
101This (rarely used) callback is called before a new connection is
102attempted, but after the file handle has been created. It could be used to
103prepare the file handle with parameters required for the actual connect
104(as opposed to settings that can be changed when the connection is already
105established).
106
107The 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
109timeout is to be used).
110
111=item on_connect => $cb->($handle, $host, $port, $retry->())
112
113This callback is called when a connection has been successfully established.
114
115The actual numeric host and port (the socket peername) are passed as
116parameters, together with a retry callback.
117
118When, for some reason, the handle is not acceptable, then calling
119C<$retry> will continue with the next conenction target (in case of
120multi-homed hosts or SRV records there can be multiple connection
121endpoints). When it is called then the read and write queues, eof status,
122tls status and similar properties of the handle are being reset.
123
124In most cases, ignoring the C<$retry> parameter is the way to go.
125
126=item on_connect_error => $cb->($handle, $message)
127
128This callback is called when the conenction could not be
129established. C<$!> will contain the relevant error code, and C<$message> a
130message describing it (usually the same as C<"$!">).
131
132If this callback isn't specified, then C<on_error> will be called with a
133fatal error instead.
134
135=back
136
137=item on_error => $cb->($handle, $fatal, $message)
138
87This is the fatal error callback, that is called when, well, a fatal error 139This is the error callback, which is called when, well, some error
88occurs, such as not being able to resolve the hostname, failure to connect 140occured, such as not being able to resolve the hostname, failure to
89or a read error. 141connect or a read error.
90 142
91The object will not be in a usable state when this callback has been 143Some errors are fatal (which is indicated by C<$fatal> being true). On
92called. 144fatal errors the handle object will be destroyed (by a call to C<< ->
145destroy >>) after invoking the error callback (which means you are free to
146examine the handle object). Examples of fatal errors are an EOF condition
147with active (but unsatisifable) read watchers (C<EPIPE>) or I/O errors. In
148cases where the other side can close the connection at their will it is
149often easiest to not report C<EPIPE> errors in this callback.
150
151AnyEvent::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
153recommended to always output the C<$message> argument in human-readable
154error messages (it's usually the same as C<"$!">).
155
156Non-fatal errors can be retried by simply returning, but it is recommended
157to simply ignore this parameter and instead abondon the handle object
158when this callback is invoked. Examples of non-fatal errors are timeouts
159C<ETIMEDOUT>) or badly-formatted data (C<EBADMSG>).
93 160
94On callback entrance, the value of C<$!> contains the operating system 161On callback entrance, the value of C<$!> contains the operating system
95error (or C<ENOSPC> or C<EPIPE>). 162error code (or C<ENOSPC>, C<EPIPE>, C<ETIMEDOUT>, C<EBADMSG> or
163C<EPROTO>).
96 164
97While not mandatory, it is I<highly> recommended to set this callback, as 165While not mandatory, it is I<highly> recommended to set this callback, as
98you will not be notified of errors otherwise. The default simply calls 166you will not be notified of errors otherwise. The default simply calls
99die. 167C<croak>.
100 168
101=item on_read => $cb->($self) 169=item on_read => $cb->($handle)
102 170
103This sets the default read callback, which is called when data arrives 171This sets the default read callback, which is called when data arrives
104and no read request is in the queue. 172and no read request is in the queue (unlike read queue callbacks, this
173callback will only be called when at least one octet of data is in the
174read buffer).
105 175
106To access (and remove data from) the read buffer, use the C<< ->rbuf >> 176To access (and remove data from) the read buffer, use the C<< ->rbuf >>
107method or access the C<$self->{rbuf}> member directly. 177method or access the C<< $handle->{rbuf} >> member directly. Note that you
178must not enlarge or modify the read buffer, you can only remove data at
179the beginning from it.
108 180
109When an EOF condition is detected then AnyEvent::Handle will first try to 181When an EOF condition is detected then AnyEvent::Handle will first try to
110feed all the remaining data to the queued callbacks and C<on_read> before 182feed all the remaining data to the queued callbacks and C<on_read> before
111calling the C<on_eof> callback. If no progress can be made, then a fatal 183calling the C<on_eof> callback. If no progress can be made, then a fatal
112error will be raised (with C<$!> set to C<EPIPE>). 184error will be raised (with C<$!> set to C<EPIPE>).
113 185
186Note that, unlike requests in the read queue, an C<on_read> callback
187doesn't mean you I<require> some data: if there is an EOF and there
188are outstanding read requests then an error will be flagged. With an
189C<on_read> callback, the C<on_eof> callback will be invoked.
190
191=item on_eof => $cb->($handle)
192
193Set the callback to be called when an end-of-file condition is detected,
194i.e. in the case of a socket, when the other side has closed the
195connection cleanly, and there are no outstanding read requests in the
196queue (if there are read requests, then an EOF counts as an unexpected
197connection close and will be flagged as an error).
198
199For sockets, this just means that the other side has stopped sending data,
200you can still try to write data, and, in fact, one can return from the EOF
201callback and continue writing data, as only the read part has been shut
202down.
203
204If an EOF condition has been detected but no C<on_eof> callback has been
205set, then a fatal error will be raised with C<$!> set to <0>.
206
114=item on_drain => $cb->() 207=item on_drain => $cb->($handle)
115 208
116This sets the callback that is called when the write buffer becomes empty 209This sets the callback that is called when the write buffer becomes empty
117(or when the callback is set and the buffer is empty already). 210(or when the callback is set and the buffer is empty already).
118 211
119To append to the write buffer, use the C<< ->push_write >> method. 212To append to the write buffer, use the C<< ->push_write >> method.
120 213
214This 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
216of some file to the socket you might not want to read the whole file into
217memory and push it into the queue, but instead only read more data from
218the file when the write queue becomes empty.
219
220=item timeout => $fractional_seconds
221
222=item rtimeout => $fractional_seconds
223
224=item wtimeout => $fractional_seconds
225
226If non-zero, then these enables an "inactivity" timeout: whenever this
227many 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
229will be invoked (and if that one is missing, a non-fatal C<ETIMEDOUT>
230error will be raised).
231
232There are three variants of the timeouts that work fully independent
233of each other, for both read and write, just read, and just write:
234C<timeout>, C<rtimeout> and C<wtimeout>, with corresponding callbacks
235C<on_timeout>, C<on_rtimeout> and C<on_wtimeout>, and reset functions
236C<timeout_reset>, C<rtimeout_reset>, and C<wtimeout_reset>.
237
238Note that timeout processing is also active when you currently do not have
239any outstanding read or write requests: If you plan to keep the connection
240idle then you should disable the timout temporarily or ignore the timeout
241in the C<on_timeout> callback, in which case AnyEvent::Handle will simply
242restart the timeout.
243
244Zero (the default) disables this timeout.
245
246=item on_timeout => $cb->($handle)
247
248Called whenever the inactivity timeout passes. If you return from this
249callback, then the timeout will be reset as if some activity had happened,
250so this condition is not fatal in any way.
251
121=item rbuf_max => <bytes> 252=item rbuf_max => <bytes>
122 253
123If defined, then a fatal error will be raised (with C<$!> set to C<ENOSPC>) 254If defined, then a fatal error will be raised (with C<$!> set to C<ENOSPC>)
124when the read buffer ever (strictly) exceeds this size. This is useful to 255when the read buffer ever (strictly) exceeds this size. This is useful to
125avoid denial-of-service attacks. 256avoid some forms of denial-of-service attacks.
126 257
127For example, a server accepting connections from untrusted sources should 258For example, a server accepting connections from untrusted sources should
128be configured to accept only so-and-so much data that it cannot act on 259be configured to accept only so-and-so much data that it cannot act on
129(for example, when expecting a line, an attacker could send an unlimited 260(for example, when expecting a line, an attacker could send an unlimited
130amount of data without a callback ever being called as long as the line 261amount of data without a callback ever being called as long as the line
131isn't finished). 262isn't finished).
132 263
264=item autocork => <boolean>
265
266When disabled (the default), then C<push_write> will try to immediately
267write 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
269be inefficient if you write multiple small chunks (on the wire, this
270disadvantage is usually avoided by your kernel's nagle algorithm, see
271C<no_delay>, but this option can save costly syscalls).
272
273When enabled, then writes will always be queued till the next event loop
274iteration. 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
276the write buffer often is full). It also increases write latency.
277
278=item no_delay => <boolean>
279
280When doing small writes on sockets, your operating system kernel might
281wait a bit for more data before actually sending it out. This is called
282the Nagle algorithm, and usually it is beneficial.
283
284In some situations you want as low a delay as possible, which can be
285accomplishd by setting this option to a true value.
286
287The default is your opertaing system's default behaviour (most likely
288enabled), this option explicitly enables or disables it, if possible.
289
290=item keepalive => <boolean>
291
292Enables (default disable) the SO_KEEPALIVE option on the stream socket:
293normally, TCP connections have no time-out once established, so TCP
294conenctions, once established, can stay alive forever even when the other
295side has long gone. TCP keepalives are a cheap way to take down long-lived
296TCP connections whent he other side becomes unreachable. While the default
297is OS-dependent, TCP keepalives usually kick in after around two hours,
298and, if the other side doesn't reply, take down the TCP connection some 10
299to 15 minutes later.
300
301It is harmless to specify this option for file handles that do not support
302keepalives, and enabling it on connections that are potentially long-lived
303is usually a good idea.
304
305=item oobinline => <boolean>
306
307BSD majorly fucked up the implementation of TCP urgent data. The result
308is that almost no OS implements TCP according to the specs, and every OS
309implements it slightly differently.
310
311If you want to handle TCP urgent data, then setting this flag gives you
312the most portable way of getting urgent data, by putting it into the
313stream.
314
133=item read_size => <bytes> 315=item read_size => <bytes>
134 316
135The default read block size (the amount of bytes this module will try to read 317The default read block size (the amount of bytes this module will
136on each [loop iteration). Default: C<4096>. 318try to read during each loop iteration, which affects memory
319requirements). Default: C<8192>.
137 320
138=item low_water_mark => <bytes> 321=item low_water_mark => <bytes>
139 322
140Sets the amount of bytes (default: C<0>) that make up an "empty" write 323Sets the amount of bytes (default: C<0>) that make up an "empty" write
141buffer: If the write reaches this size or gets even samller it is 324buffer: If the write reaches this size or gets even samller it is
142considered empty. 325considered empty.
143 326
327Sometimes it can be beneficial (for performance reasons) to add data to
328the write buffer before it is fully drained, but this is a rare case, as
329the operating system kernel usually buffers data as well, so the default
330is good in almost all cases.
331
332=item linger => <seconds>
333
334If non-zero (default: C<3600>), then the destructor of the
335AnyEvent::Handle object will check whether there is still outstanding
336write data and will install a watcher that will write this data to the
337socket. No errors will be reported (this mostly matches how the operating
338system treats outstanding data at socket close time).
339
340This will not work for partial TLS data that could not be encoded
341yet. This data will be lost. Calling the C<stoptls> method in time might
342help.
343
344=item peername => $string
345
346A string used to identify the remote site - usually the DNS hostname
347(I<not> IDN!) used to create the connection, rarely the IP address.
348
349Apart from being useful in error messages, this string is also used in TLS
350peername verification (see C<verify_peername> in L<AnyEvent::TLS>). This
351verification will be skipped when C<peername> is not specified or
352C<undef>.
353
144=item tls => "accept" | "connect" | Net::SSLeay::SSL object 354=item tls => "accept" | "connect" | Net::SSLeay::SSL object
145 355
146When this parameter is given, it enables TLS (SSL) mode, that means it 356When this parameter is given, it enables TLS (SSL) mode, that means
147will start making tls handshake and will transparently encrypt/decrypt 357AnyEvent will start a TLS handshake as soon as the conenction has been
148data. 358established and will transparently encrypt/decrypt data afterwards.
149 359
150For the TLS server side, use C<accept>, and for the TLS client side of a 360All TLS protocol errors will be signalled as C<EPROTO>, with an
151connection, use C<connect> mode. 361appropriate error message.
362
363TLS mode requires Net::SSLeay to be installed (it will be loaded
364automatically when you try to create a TLS handle): this module doesn't
365have a dependency on that module, so if your module requires it, you have
366to add the dependency yourself.
367
368Unlike TCP, TLS has a server and client side: for the TLS server side, use
369C<accept>, and for the TLS client side of a connection, use C<connect>
370mode.
152 371
153You can also provide your own TLS connection object, but you have 372You can also provide your own TLS connection object, but you have
154to make sure that you call either C<Net::SSLeay::set_connect_state> 373to make sure that you call either C<Net::SSLeay::set_connect_state>
155or C<Net::SSLeay::set_accept_state> on it before you pass it to 374or C<Net::SSLeay::set_accept_state> on it before you pass it to
156AnyEvent::Handle. 375AnyEvent::Handle. Also, this module will take ownership of this connection
376object.
157 377
158=item tls_ctx => $ssl_ctx 378At some future point, AnyEvent::Handle might switch to another TLS
379implementation, then the option to use your own session object will go
380away.
159 381
382B<IMPORTANT:> since Net::SSLeay "objects" are really only integers,
383passing in the wrong integer will lead to certain crash. This most often
384happens when one uses a stylish C<< tls => 1 >> and is surprised about the
385segmentation fault.
386
387See the C<< ->starttls >> method for when need to start TLS negotiation later.
388
389=item tls_ctx => $anyevent_tls
390
160Use the given Net::SSLeay::CTX object to create the new TLS connection 391Use the given C<AnyEvent::TLS> object to create the new TLS connection
161(unless a connection object was specified directly). If this parameter is 392(unless a connection object was specified directly). If this parameter is
162missing, then AnyEvent::Handle will use C<AnyEvent::Handle::TLS_CTX>. 393missing, then AnyEvent::Handle will use C<AnyEvent::Handle::TLS_CTX>.
163 394
395Instead of an object, you can also specify a hash reference with C<< key
396=> value >> pairs. Those will be passed to L<AnyEvent::TLS> to create a
397new TLS context object.
398
399=item on_starttls => $cb->($handle, $success[, $error_message])
400
401This callback will be invoked when the TLS/SSL handshake has finished. If
402C<$success> is true, then the TLS handshake succeeded, otherwise it failed
403(C<on_stoptls> will not be called in this case).
404
405The session in C<< $handle->{tls} >> can still be examined in this
406callback, even when the handshake was not successful.
407
408TLS handshake failures will not cause C<on_error> to be invoked when this
409callback is in effect, instead, the error message will be passed to C<on_starttls>.
410
411Without this callback, handshake failures lead to C<on_error> being
412called, as normal.
413
414Note that you cannot call C<starttls> right again in this callback. If you
415need to do that, start an zero-second timer instead whose callback can
416then call C<< ->starttls >> again.
417
418=item on_stoptls => $cb->($handle)
419
420When a SSLv3/TLS shutdown/close notify/EOF is detected and this callback is
421set, then it will be invoked after freeing the TLS session. If it is not,
422then a TLS shutdown condition will be treated like a normal EOF condition
423on the handle.
424
425The session in C<< $handle->{tls} >> can still be examined in this
426callback.
427
428This callback will only be called on TLS shutdowns, not when the
429underlying handle signals EOF.
430
431=item json => JSON or JSON::XS object
432
433This is the json coder object used by the C<json> read and write types.
434
435If you don't supply it, then AnyEvent::Handle will create and use a
436suitable one (on demand), which will write and expect UTF-8 encoded JSON
437texts.
438
439Note that you are responsible to depend on the JSON module if you want to
440use this functionality, as AnyEvent does not have a dependency itself.
441
164=back 442=back
165 443
166=cut 444=cut
167 445
168sub new { 446sub new {
169 my $class = shift; 447 my $class = shift;
170
171 my $self = bless { @_ }, $class; 448 my $self = bless { @_ }, $class;
172 449
173 $self->{fh} or Carp::croak "mandatory argument fh is missing"; 450 if ($self->{fh}) {
451 $self->_start;
452 return unless $self->{fh}; # could be gone by now
453
454 } elsif ($self->{connect}) {
455 require AnyEvent::Socket;
456
457 $self->{peername} = $self->{connect}[0]
458 unless exists $self->{peername};
459
460 $self->{_skip_drain_rbuf} = 1;
461
462 {
463 Scalar::Util::weaken (my $self = $self);
464
465 $self->{_connect} =
466 AnyEvent::Socket::tcp_connect (
467 $self->{connect}[0],
468 $self->{connect}[1],
469 sub {
470 my ($fh, $host, $port, $retry) = @_;
471
472 if ($fh) {
473 $self->{fh} = $fh;
474
475 delete $self->{_skip_drain_rbuf};
476 $self->_start;
477
478 $self->{on_connect}
479 and $self->{on_connect}($self, $host, $port, sub {
480 delete @$self{qw(fh _tw _rtw _wtw _ww _rw _eof _queue rbuf _wbuf tls _tls_rbuf _tls_wbuf)};
481 $self->{_skip_drain_rbuf} = 1;
482 &$retry;
483 });
484
485 } else {
486 if ($self->{on_connect_error}) {
487 $self->{on_connect_error}($self, "$!");
488 $self->destroy;
489 } else {
490 $self->_error ($!, 1);
491 }
492 }
493 },
494 sub {
495 local $self->{fh} = $_[0];
496
497 $self->{on_prepare}
498 ? $self->{on_prepare}->($self)
499 : ()
500 }
501 );
502 }
503
504 } else {
505 Carp::croak "AnyEvent::Handle: either an existing fh or the connect parameter must be specified";
506 }
507
508 $self
509}
510
511sub _start {
512 my ($self) = @_;
174 513
175 AnyEvent::Util::fh_nonblocking $self->{fh}, 1; 514 AnyEvent::Util::fh_nonblocking $self->{fh}, 1;
176 515
177 if ($self->{tls}) { 516 $self->{_activity} =
178 require Net::SSLeay; 517 $self->{_ractivity} =
518 $self->{_wactivity} = AE::now;
519
520 $self->timeout (delete $self->{timeout} ) if $self->{timeout};
521 $self->rtimeout (delete $self->{rtimeout} ) if $self->{rtimeout};
522 $self->wtimeout (delete $self->{wtimeout} ) if $self->{wtimeout};
523
524 $self->no_delay (delete $self->{no_delay} ) if exists $self->{no_delay};
525 $self->keepalive (delete $self->{keepalive}) if exists $self->{keepalive};
526 $self->oobinline (delete $self->{oobinline}) if exists $self->{oobinline};
527
179 $self->starttls (delete $self->{tls}, delete $self->{tls_ctx}); 528 $self->starttls (delete $self->{tls}, delete $self->{tls_ctx})
180 } 529 if $self->{tls};
181 530
182 $self->on_eof (delete $self->{on_eof} ) if $self->{on_eof};
183 $self->on_error (delete $self->{on_error}) if $self->{on_error};
184 $self->on_drain (delete $self->{on_drain}) if $self->{on_drain}; 531 $self->on_drain (delete $self->{on_drain}) if $self->{on_drain};
185 $self->on_read (delete $self->{on_read} ) if $self->{on_read};
186 532
187 $self->start_read; 533 $self->start_read
534 if $self->{on_read} || @{ $self->{_queue} };
188 535
189 $self 536 $self->_drain_wbuf;
190} 537}
191 538
192sub _shutdown {
193 my ($self) = @_;
194
195 delete $self->{rw};
196 delete $self->{ww};
197 delete $self->{fh};
198}
199
200sub error { 539sub _error {
201 my ($self) = @_; 540 my ($self, $errno, $fatal, $message) = @_;
202 541
203 { 542 $! = $errno;
204 local $!; 543 $message ||= "$!";
205 $self->_shutdown;
206 }
207 544
208 if ($self->{on_error}) { 545 if ($self->{on_error}) {
209 $self->{on_error}($self); 546 $self->{on_error}($self, $fatal, $message);
210 } else { 547 $self->destroy if $fatal;
548 } elsif ($self->{fh}) {
549 $self->destroy;
211 die "AnyEvent::Handle uncaught fatal error: $!"; 550 Carp::croak "AnyEvent::Handle uncaught error: $message";
212 } 551 }
213} 552}
214 553
215=item $fh = $handle->fh 554=item $fh = $handle->fh
216 555
217This method returns the file handle of the L<AnyEvent::Handle> object. 556This method returns the file handle used to create the L<AnyEvent::Handle> object.
218 557
219=cut 558=cut
220 559
221sub fh { $_[0]->{fh} } 560sub fh { $_[0]{fh} }
222 561
223=item $handle->on_error ($cb) 562=item $handle->on_error ($cb)
224 563
225Replace the current C<on_error> callback (see the C<on_error> constructor argument). 564Replace the current C<on_error> callback (see the C<on_error> constructor argument).
226 565
238 577
239sub on_eof { 578sub on_eof {
240 $_[0]{on_eof} = $_[1]; 579 $_[0]{on_eof} = $_[1];
241} 580}
242 581
582=item $handle->on_timeout ($cb)
583
584=item $handle->on_rtimeout ($cb)
585
586=item $handle->on_wtimeout ($cb)
587
588Replace the current C<on_timeout>, C<on_rtimeout> or C<on_wtimeout>
589callback, or disables the callback (but not the timeout) if C<$cb> =
590C<undef>. See the C<timeout> constructor argument and method.
591
592=cut
593
594# see below
595
596=item $handle->autocork ($boolean)
597
598Enables or disables the current autocork behaviour (see C<autocork>
599constructor argument). Changes will only take effect on the next write.
600
601=cut
602
603sub autocork {
604 $_[0]{autocork} = $_[1];
605}
606
607=item $handle->no_delay ($boolean)
608
609Enables or disables the C<no_delay> setting (see constructor argument of
610the same name for details).
611
612=cut
613
614sub no_delay {
615 $_[0]{no_delay} = $_[1];
616
617 eval {
618 local $SIG{__DIE__};
619 setsockopt $_[0]{fh}, Socket::IPPROTO_TCP (), Socket::TCP_NODELAY (), int $_[1]
620 if $_[0]{fh};
621 };
622}
623
624=item $handle->keepalive ($boolean)
625
626Enables or disables the C<keepalive> setting (see constructor argument of
627the same name for details).
628
629=cut
630
631sub keepalive {
632 $_[0]{keepalive} = $_[1];
633
634 eval {
635 local $SIG{__DIE__};
636 setsockopt $_[0]{fh}, Socket::SOL_SOCKET (), Socket::SO_KEEPALIVE (), int $_[1]
637 if $_[0]{fh};
638 };
639}
640
641=item $handle->oobinline ($boolean)
642
643Enables or disables the C<oobinline> setting (see constructor argument of
644the same name for details).
645
646=cut
647
648sub oobinline {
649 $_[0]{oobinline} = $_[1];
650
651 eval {
652 local $SIG{__DIE__};
653 setsockopt $_[0]{fh}, Socket::SOL_SOCKET (), Socket::SO_OOBINLINE (), int $_[1]
654 if $_[0]{fh};
655 };
656}
657
658=item $handle->keepalive ($boolean)
659
660Enables or disables the C<keepalive> setting (see constructor argument of
661the same name for details).
662
663=cut
664
665sub keepalive {
666 $_[0]{keepalive} = $_[1];
667
668 eval {
669 local $SIG{__DIE__};
670 setsockopt $_[0]{fh}, Socket::SOL_SOCKET (), Socket::SO_KEEPALIVE (), int $_[1]
671 if $_[0]{fh};
672 };
673}
674
675=item $handle->on_starttls ($cb)
676
677Replace the current C<on_starttls> callback (see the C<on_starttls> constructor argument).
678
679=cut
680
681sub on_starttls {
682 $_[0]{on_starttls} = $_[1];
683}
684
685=item $handle->on_stoptls ($cb)
686
687Replace the current C<on_stoptls> callback (see the C<on_stoptls> constructor argument).
688
689=cut
690
691sub on_starttls {
692 $_[0]{on_stoptls} = $_[1];
693}
694
695=item $handle->rbuf_max ($max_octets)
696
697Configures the C<rbuf_max> setting (C<undef> disables it).
698
699=cut
700
701sub rbuf_max {
702 $_[0]{rbuf_max} = $_[1];
703}
704
705#############################################################################
706
707=item $handle->timeout ($seconds)
708
709=item $handle->rtimeout ($seconds)
710
711=item $handle->wtimeout ($seconds)
712
713Configures (or disables) the inactivity timeout.
714
715=item $handle->timeout_reset
716
717=item $handle->rtimeout_reset
718
719=item $handle->wtimeout_reset
720
721Reset the activity timeout, as if data was received or sent.
722
723These methods are cheap to call.
724
725=cut
726
727for my $dir ("", "r", "w") {
728 my $timeout = "${dir}timeout";
729 my $tw = "_${dir}tw";
730 my $on_timeout = "on_${dir}timeout";
731 my $activity = "_${dir}activity";
732 my $cb;
733
734 *$on_timeout = sub {
735 $_[0]{$on_timeout} = $_[1];
736 };
737
738 *$timeout = sub {
739 my ($self, $new_value) = @_;
740
741 $self->{$timeout} = $new_value;
742 delete $self->{$tw}; &$cb;
743 };
744
745 *{"${dir}timeout_reset"} = sub {
746 $_[0]{$activity} = AE::now;
747 };
748
749 # main workhorse:
750 # reset the timeout watcher, as neccessary
751 # also check for time-outs
752 $cb = sub {
753 my ($self) = @_;
754
755 if ($self->{$timeout} && $self->{fh}) {
756 my $NOW = AE::now;
757
758 # when would the timeout trigger?
759 my $after = $self->{$activity} + $self->{$timeout} - $NOW;
760
761 # now or in the past already?
762 if ($after <= 0) {
763 $self->{$activity} = $NOW;
764
765 if ($self->{$on_timeout}) {
766 $self->{$on_timeout}($self);
767 } else {
768 $self->_error (Errno::ETIMEDOUT);
769 }
770
771 # callback could have changed timeout value, optimise
772 return unless $self->{$timeout};
773
774 # calculate new after
775 $after = $self->{$timeout};
776 }
777
778 Scalar::Util::weaken $self;
779 return unless $self; # ->error could have destroyed $self
780
781 $self->{$tw} ||= AE::timer $after, 0, sub {
782 delete $self->{$tw};
783 $cb->($self);
784 };
785 } else {
786 delete $self->{$tw};
787 }
788 }
789}
790
243############################################################################# 791#############################################################################
244 792
245=back 793=back
246 794
247=head2 WRITE QUEUE 795=head2 WRITE QUEUE
268 my ($self, $cb) = @_; 816 my ($self, $cb) = @_;
269 817
270 $self->{on_drain} = $cb; 818 $self->{on_drain} = $cb;
271 819
272 $cb->($self) 820 $cb->($self)
273 if $cb && $self->{low_water_mark} >= length $self->{wbuf}; 821 if $cb && $self->{low_water_mark} >= (length $self->{wbuf}) + (length $self->{_tls_wbuf});
274} 822}
275 823
276=item $handle->push_write ($data) 824=item $handle->push_write ($data)
277 825
278Queues the given scalar to be written. You can push as much data as you 826Queues the given scalar to be written. You can push as much data as you
282=cut 830=cut
283 831
284sub _drain_wbuf { 832sub _drain_wbuf {
285 my ($self) = @_; 833 my ($self) = @_;
286 834
287 unless ($self->{ww}) { 835 if (!$self->{_ww} && length $self->{wbuf}) {
836
288 Scalar::Util::weaken $self; 837 Scalar::Util::weaken $self;
838
289 my $cb = sub { 839 my $cb = sub {
290 my $len = syswrite $self->{fh}, $self->{wbuf}; 840 my $len = syswrite $self->{fh}, $self->{wbuf};
291 841
292 if ($len > 0) { 842 if (defined $len) {
293 substr $self->{wbuf}, 0, $len, ""; 843 substr $self->{wbuf}, 0, $len, "";
294 844
845 $self->{_activity} = $self->{_wactivity} = AE::now;
846
295 $self->{on_drain}($self) 847 $self->{on_drain}($self)
296 if $self->{low_water_mark} >= length $self->{wbuf} 848 if $self->{low_water_mark} >= (length $self->{wbuf}) + (length $self->{_tls_wbuf})
297 && $self->{on_drain}; 849 && $self->{on_drain};
298 850
299 delete $self->{ww} unless length $self->{wbuf}; 851 delete $self->{_ww} unless length $self->{wbuf};
300 } elsif ($! != EAGAIN && $! != EINTR) { 852 } elsif ($! != EAGAIN && $! != EINTR && $! != WSAEWOULDBLOCK) {
301 $self->error; 853 $self->_error ($!, 1);
302 } 854 }
303 }; 855 };
304 856
305 $self->{ww} = AnyEvent->io (fh => $self->{fh}, poll => "w", cb => $cb); 857 # try to write data immediately
858 $cb->() unless $self->{autocork};
306 859
307 $cb->($self); 860 # if still data left in wbuf, we need to poll
861 $self->{_ww} = AE::io $self->{fh}, 1, $cb
862 if length $self->{wbuf};
308 }; 863 };
864}
865
866our %WH;
867
868sub register_write_type($$) {
869 $WH{$_[0]} = $_[1];
309} 870}
310 871
311sub push_write { 872sub push_write {
312 my $self = shift; 873 my $self = shift;
313 874
875 if (@_ > 1) {
876 my $type = shift;
877
878 @_ = ($WH{$type} or Carp::croak "unsupported type passed to AnyEvent::Handle::push_write")
879 ->($self, @_);
880 }
881
314 if ($self->{filter_w}) { 882 if ($self->{tls}) {
315 $self->{filter_w}->($self, \$_[0]); 883 $self->{_tls_wbuf} .= $_[0];
884 &_dotls ($self) if $self->{fh};
316 } else { 885 } else {
317 $self->{wbuf} .= $_[0]; 886 $self->{wbuf} .= $_[0];
318 $self->_drain_wbuf; 887 $self->_drain_wbuf if $self->{fh};
319 } 888 }
320} 889}
890
891=item $handle->push_write (type => @args)
892
893Instead of formatting your data yourself, you can also let this module do
894the job by specifying a type and type-specific arguments.
895
896Predefined types are (if you have ideas for additional types, feel free to
897drop by and tell us):
898
899=over 4
900
901=item netstring => $string
902
903Formats the given value as netstring
904(http://cr.yp.to/proto/netstrings.txt, this is not a recommendation to use them).
905
906=cut
907
908register_write_type netstring => sub {
909 my ($self, $string) = @_;
910
911 (length $string) . ":$string,"
912};
913
914=item packstring => $format, $data
915
916An octet string prefixed with an encoded length. The encoding C<$format>
917uses the same format as a Perl C<pack> format, but must specify a single
918integer only (only one of C<cCsSlLqQiInNvVjJw> is allowed, plus an
919optional C<!>, C<< < >> or C<< > >> modifier).
920
921=cut
922
923register_write_type packstring => sub {
924 my ($self, $format, $string) = @_;
925
926 pack "$format/a*", $string
927};
928
929=item json => $array_or_hashref
930
931Encodes the given hash or array reference into a JSON object. Unless you
932provide your own JSON object, this means it will be encoded to JSON text
933in UTF-8.
934
935JSON objects (and arrays) are self-delimiting, so you can write JSON at
936one end of a handle and read them at the other end without using any
937additional framing.
938
939The generated JSON text is guaranteed not to contain any newlines: While
940this module doesn't need delimiters after or between JSON texts to be
941able to read them, many other languages depend on that.
942
943A simple RPC protocol that interoperates easily with others is to send
944JSON arrays (or objects, although arrays are usually the better choice as
945they mimic how function argument passing works) and a newline after each
946JSON text:
947
948 $handle->push_write (json => ["method", "arg1", "arg2"]); # whatever
949 $handle->push_write ("\012");
950
951An AnyEvent::Handle receiver would simply use the C<json> read type and
952rely on the fact that the newline will be skipped as leading whitespace:
953
954 $handle->push_read (json => sub { my $array = $_[1]; ... });
955
956Other languages could read single lines terminated by a newline and pass
957this line into their JSON decoder of choice.
958
959=cut
960
961sub json_coder() {
962 eval { require JSON::XS; JSON::XS->new->utf8 }
963 || do { require JSON; JSON->new->utf8 }
964}
965
966register_write_type json => sub {
967 my ($self, $ref) = @_;
968
969 my $json = $self->{json} ||= json_coder;
970
971 $json->encode ($ref)
972};
973
974=item storable => $reference
975
976Freezes the given reference using L<Storable> and writes it to the
977handle. Uses the C<nfreeze> format.
978
979=cut
980
981register_write_type storable => sub {
982 my ($self, $ref) = @_;
983
984 require Storable;
985
986 pack "w/a*", Storable::nfreeze ($ref)
987};
988
989=back
990
991=item $handle->push_shutdown
992
993Sometimes you know you want to close the socket after writing your data
994before it was actually written. One way to do that is to replace your
995C<on_drain> handler by a callback that shuts down the socket (and set
996C<low_water_mark> to C<0>). This method is a shorthand for just that, and
997replaces the C<on_drain> callback with:
998
999 sub { shutdown $_[0]{fh}, 1 } # for push_shutdown
1000
1001This simply shuts down the write side and signals an EOF condition to the
1002the peer.
1003
1004You can rely on the normal read queue and C<on_eof> handling
1005afterwards. This is the cleanest way to close a connection.
1006
1007=cut
1008
1009sub push_shutdown {
1010 my ($self) = @_;
1011
1012 delete $self->{low_water_mark};
1013 $self->on_drain (sub { shutdown $_[0]{fh}, 1 });
1014}
1015
1016=item AnyEvent::Handle::register_write_type type => $coderef->($handle, @args)
1017
1018This function (not method) lets you add your own types to C<push_write>.
1019Whenever the given C<type> is used, C<push_write> will invoke the code
1020reference with the handle object and the remaining arguments.
1021
1022The code reference is supposed to return a single octet string that will
1023be appended to the write buffer.
1024
1025Note that this is a function, and all types registered this way will be
1026global, so try to use unique names.
1027
1028=cut
321 1029
322############################################################################# 1030#############################################################################
323 1031
324=back 1032=back
325 1033
332ways, the "simple" way, using only C<on_read> and the "complex" way, using 1040ways, the "simple" way, using only C<on_read> and the "complex" way, using
333a queue. 1041a queue.
334 1042
335In the simple case, you just install an C<on_read> callback and whenever 1043In the simple case, you just install an C<on_read> callback and whenever
336new data arrives, it will be called. You can then remove some data (if 1044new data arrives, it will be called. You can then remove some data (if
337enough is there) from the read buffer (C<< $handle->rbuf >>) if you want 1045enough is there) from the read buffer (C<< $handle->rbuf >>). Or you cna
338or not. 1046leave the data there if you want to accumulate more (e.g. when only a
1047partial message has been received so far).
339 1048
340In the more complex case, you want to queue multiple callbacks. In this 1049In the more complex case, you want to queue multiple callbacks. In this
341case, AnyEvent::Handle will call the first queued callback each time new 1050case, AnyEvent::Handle will call the first queued callback each time new
342data arrives and removes it when it has done its job (see C<push_read>, 1051data arrives (also the first time it is queued) and removes it when it has
343below). 1052done its job (see C<push_read>, below).
344 1053
345This way you can, for example, push three line-reads, followed by reading 1054This way you can, for example, push three line-reads, followed by reading
346a chunk of data, and AnyEvent::Handle will execute them in order. 1055a chunk of data, and AnyEvent::Handle will execute them in order.
347 1056
348Example 1: EPP protocol parser. EPP sends 4 byte length info, followed by 1057Example 1: EPP protocol parser. EPP sends 4 byte length info, followed by
349the specified number of bytes which give an XML datagram. 1058the specified number of bytes which give an XML datagram.
350 1059
351 # in the default state, expect some header bytes 1060 # in the default state, expect some header bytes
352 $handle->on_read (sub { 1061 $handle->on_read (sub {
353 # some data is here, now queue the length-header-read (4 octets) 1062 # some data is here, now queue the length-header-read (4 octets)
354 shift->unshift_read_chunk (4, sub { 1063 shift->unshift_read (chunk => 4, sub {
355 # header arrived, decode 1064 # header arrived, decode
356 my $len = unpack "N", $_[1]; 1065 my $len = unpack "N", $_[1];
357 1066
358 # now read the payload 1067 # now read the payload
359 shift->unshift_read_chunk ($len, sub { 1068 shift->unshift_read (chunk => $len, sub {
360 my $xml = $_[1]; 1069 my $xml = $_[1];
361 # handle xml 1070 # handle xml
362 }); 1071 });
363 }); 1072 });
364 }); 1073 });
365 1074
366Example 2: Implement a client for a protocol that replies either with 1075Example 2: Implement a client for a protocol that replies either with "OK"
367"OK" and another line or "ERROR" for one request, and 64 bytes for the 1076and another line or "ERROR" for the first request that is sent, and 64
368second request. Due tot he availability of a full queue, we can just 1077bytes for the second request. Due to the availability of a queue, we can
369pipeline sending both requests and manipulate the queue as necessary in 1078just pipeline sending both requests and manipulate the queue as necessary
370the callbacks: 1079in the callbacks.
371 1080
372 # request one 1081When the first callback is called and sees an "OK" response, it will
1082C<unshift> another line-read. This line-read will be queued I<before> the
108364-byte chunk callback.
1084
1085 # request one, returns either "OK + extra line" or "ERROR"
373 $handle->push_write ("request 1\015\012"); 1086 $handle->push_write ("request 1\015\012");
374 1087
375 # we expect "ERROR" or "OK" as response, so push a line read 1088 # we expect "ERROR" or "OK" as response, so push a line read
376 $handle->push_read_line (sub { 1089 $handle->push_read (line => sub {
377 # if we got an "OK", we have to _prepend_ another line, 1090 # if we got an "OK", we have to _prepend_ another line,
378 # so it will be read before the second request reads its 64 bytes 1091 # so it will be read before the second request reads its 64 bytes
379 # which are already in the queue when this callback is called 1092 # which are already in the queue when this callback is called
380 # we don't do this in case we got an error 1093 # we don't do this in case we got an error
381 if ($_[1] eq "OK") { 1094 if ($_[1] eq "OK") {
382 $_[0]->unshift_read_line (sub { 1095 $_[0]->unshift_read (line => sub {
383 my $response = $_[1]; 1096 my $response = $_[1];
384 ... 1097 ...
385 }); 1098 });
386 } 1099 }
387 }); 1100 });
388 1101
389 # request two 1102 # request two, simply returns 64 octets
390 $handle->push_write ("request 2\015\012"); 1103 $handle->push_write ("request 2\015\012");
391 1104
392 # simply read 64 bytes, always 1105 # simply read 64 bytes, always
393 $handle->push_read_chunk (64, sub { 1106 $handle->push_read (chunk => 64, sub {
394 my $response = $_[1]; 1107 my $response = $_[1];
395 ... 1108 ...
396 }); 1109 });
397 1110
398=over 4 1111=over 4
399 1112
400=cut 1113=cut
401 1114
402sub _drain_rbuf { 1115sub _drain_rbuf {
403 my ($self) = @_; 1116 my ($self) = @_;
1117
1118 # avoid recursion
1119 return if $self->{_skip_drain_rbuf};
1120 local $self->{_skip_drain_rbuf} = 1;
1121
1122 while () {
1123 # we need to use a separate tls read buffer, as we must not receive data while
1124 # we are draining the buffer, and this can only happen with TLS.
1125 $self->{rbuf} .= delete $self->{_tls_rbuf}
1126 if exists $self->{_tls_rbuf};
1127
1128 my $len = length $self->{rbuf};
1129
1130 if (my $cb = shift @{ $self->{_queue} }) {
1131 unless ($cb->($self)) {
1132 # no progress can be made
1133 # (not enough data and no data forthcoming)
1134 $self->_error (Errno::EPIPE, 1), return
1135 if $self->{_eof};
1136
1137 unshift @{ $self->{_queue} }, $cb;
1138 last;
1139 }
1140 } elsif ($self->{on_read}) {
1141 last unless $len;
1142
1143 $self->{on_read}($self);
1144
1145 if (
1146 $len == length $self->{rbuf} # if no data has been consumed
1147 && !@{ $self->{_queue} } # and the queue is still empty
1148 && $self->{on_read} # but we still have on_read
1149 ) {
1150 # no further data will arrive
1151 # so no progress can be made
1152 $self->_error (Errno::EPIPE, 1), return
1153 if $self->{_eof};
1154
1155 last; # more data might arrive
1156 }
1157 } else {
1158 # read side becomes idle
1159 delete $self->{_rw} unless $self->{tls};
1160 last;
1161 }
1162 }
1163
1164 if ($self->{_eof}) {
1165 $self->{on_eof}
1166 ? $self->{on_eof}($self)
1167 : $self->_error (0, 1, "Unexpected end-of-file");
1168
1169 return;
1170 }
404 1171
405 if ( 1172 if (
406 defined $self->{rbuf_max} 1173 defined $self->{rbuf_max}
407 && $self->{rbuf_max} < length $self->{rbuf} 1174 && $self->{rbuf_max} < length $self->{rbuf}
408 ) { 1175 ) {
409 $! = &Errno::ENOSPC; return $self->error; 1176 $self->_error (Errno::ENOSPC, 1), return;
410 } 1177 }
411 1178
412 return if $self->{in_drain}; 1179 # may need to restart read watcher
413 local $self->{in_drain} = 1; 1180 unless ($self->{_rw}) {
414 1181 $self->start_read
415 while (my $len = length $self->{rbuf}) { 1182 if $self->{on_read} || @{ $self->{_queue} };
416 no strict 'refs';
417 if (my $cb = shift @{ $self->{queue} }) {
418 if (!$cb->($self)) {
419 if ($self->{eof}) {
420 # no progress can be made (not enough data and no data forthcoming)
421 $! = &Errno::EPIPE; return $self->error;
422 }
423
424 unshift @{ $self->{queue} }, $cb;
425 return;
426 }
427 } elsif ($self->{on_read}) {
428 $self->{on_read}($self);
429
430 if (
431 $self->{eof} # if no further data will arrive
432 && $len == length $self->{rbuf} # and no data has been consumed
433 && !@{ $self->{queue} } # and the queue is still empty
434 && $self->{on_read} # and we still want to read data
435 ) {
436 # then no progress can be made
437 $! = &Errno::EPIPE; return $self->error;
438 }
439 } else {
440 # read side becomes idle
441 delete $self->{rw};
442 return;
443 }
444 }
445
446 if ($self->{eof}) {
447 $self->_shutdown;
448 $self->{on_eof}($self)
449 if $self->{on_eof};
450 } 1183 }
451} 1184}
452 1185
453=item $handle->on_read ($cb) 1186=item $handle->on_read ($cb)
454 1187
460 1193
461sub on_read { 1194sub on_read {
462 my ($self, $cb) = @_; 1195 my ($self, $cb) = @_;
463 1196
464 $self->{on_read} = $cb; 1197 $self->{on_read} = $cb;
1198 $self->_drain_rbuf if $cb;
465} 1199}
466 1200
467=item $handle->rbuf 1201=item $handle->rbuf
468 1202
469Returns the read buffer (as a modifiable lvalue). 1203Returns the read buffer (as a modifiable lvalue).
470 1204
471You can access the read buffer directly as the C<< ->{rbuf} >> member, if 1205You can access the read buffer directly as the C<< ->{rbuf} >>
472you want. 1206member, if you want. However, the only operation allowed on the
1207read buffer (apart from looking at it) is removing data from its
1208beginning. Otherwise modifying or appending to it is not allowed and will
1209lead to hard-to-track-down bugs.
473 1210
474NOTE: The read buffer should only be used or modified if the C<on_read>, 1211NOTE: The read buffer should only be used or modified if the C<on_read>,
475C<push_read> or C<unshift_read> methods are used. The other read methods 1212C<push_read> or C<unshift_read> methods are used. The other read methods
476automatically manage the read buffer. 1213automatically manage the read buffer.
477 1214
500interested in (which can be none at all) and return a true value. After returning 1237interested in (which can be none at all) and return a true value. After returning
501true, it will be removed from the queue. 1238true, it will be removed from the queue.
502 1239
503=cut 1240=cut
504 1241
1242our %RH;
1243
1244sub register_read_type($$) {
1245 $RH{$_[0]} = $_[1];
1246}
1247
505sub push_read { 1248sub push_read {
506 my ($self, $cb) = @_; 1249 my $self = shift;
1250 my $cb = pop;
507 1251
1252 if (@_) {
1253 my $type = shift;
1254
1255 $cb = ($RH{$type} or Carp::croak "unsupported type passed to AnyEvent::Handle::push_read")
1256 ->($self, $cb, @_);
1257 }
1258
508 push @{ $self->{queue} }, $cb; 1259 push @{ $self->{_queue} }, $cb;
509 $self->_drain_rbuf; 1260 $self->_drain_rbuf;
510} 1261}
511 1262
512sub unshift_read { 1263sub unshift_read {
513 my ($self, $cb) = @_; 1264 my $self = shift;
1265 my $cb = pop;
514 1266
1267 if (@_) {
1268 my $type = shift;
1269
1270 $cb = ($RH{$type} or Carp::croak "unsupported type passed to AnyEvent::Handle::unshift_read")
1271 ->($self, $cb, @_);
1272 }
1273
515 push @{ $self->{queue} }, $cb; 1274 unshift @{ $self->{_queue} }, $cb;
516 $self->_drain_rbuf; 1275 $self->_drain_rbuf;
517} 1276}
518 1277
519=item $handle->push_read_chunk ($len, $cb->($self, $data)) 1278=item $handle->push_read (type => @args, $cb)
520 1279
521=item $handle->unshift_read_chunk ($len, $cb->($self, $data)) 1280=item $handle->unshift_read (type => @args, $cb)
522 1281
523Append the given callback to the end of the queue (C<push_read_chunk>) or 1282Instead of providing a callback that parses the data itself you can chose
524prepend it (C<unshift_read_chunk>). 1283between a number of predefined parsing formats, for chunks of data, lines
1284etc.
525 1285
526The callback will be called only once C<$len> bytes have been read, and 1286Predefined types are (if you have ideas for additional types, feel free to
527these C<$len> bytes will be passed to the callback. 1287drop by and tell us):
528 1288
529=cut 1289=over 4
530 1290
531sub _read_chunk($$) { 1291=item chunk => $octets, $cb->($handle, $data)
1292
1293Invoke the callback only once C<$octets> bytes have been read. Pass the
1294data read to the callback. The callback will never be called with less
1295data.
1296
1297Example: read 2 bytes.
1298
1299 $handle->push_read (chunk => 2, sub {
1300 warn "yay ", unpack "H*", $_[1];
1301 });
1302
1303=cut
1304
1305register_read_type chunk => sub {
532 my ($self, $len, $cb) = @_; 1306 my ($self, $cb, $len) = @_;
533 1307
534 sub { 1308 sub {
535 $len <= length $_[0]{rbuf} or return; 1309 $len <= length $_[0]{rbuf} or return;
536 $cb->($_[0], substr $_[0]{rbuf}, 0, $len, ""); 1310 $cb->($_[0], substr $_[0]{rbuf}, 0, $len, "");
537 1 1311 1
538 } 1312 }
539} 1313};
540 1314
541sub push_read_chunk { 1315=item line => [$eol, ]$cb->($handle, $line, $eol)
542 $_[0]->push_read (&_read_chunk);
543}
544
545
546sub unshift_read_chunk {
547 $_[0]->unshift_read (&_read_chunk);
548}
549
550=item $handle->push_read_line ([$eol, ]$cb->($self, $line, $eol))
551
552=item $handle->unshift_read_line ([$eol, ]$cb->($self, $line, $eol))
553
554Append the given callback to the end of the queue (C<push_read_line>) or
555prepend it (C<unshift_read_line>).
556 1316
557The callback will be called only once a full line (including the end of 1317The callback will be called only once a full line (including the end of
558line marker, C<$eol>) has been read. This line (excluding the end of line 1318line marker, C<$eol>) has been read. This line (excluding the end of line
559marker) will be passed to the callback as second argument (C<$line>), and 1319marker) will be passed to the callback as second argument (C<$line>), and
560the end of line marker as the third argument (C<$eol>). 1320the end of line marker as the third argument (C<$eol>).
571Partial lines at the end of the stream will never be returned, as they are 1331Partial lines at the end of the stream will never be returned, as they are
572not marked by the end of line marker. 1332not marked by the end of line marker.
573 1333
574=cut 1334=cut
575 1335
576sub _read_line($$) { 1336register_read_type line => sub {
577 my $self = shift; 1337 my ($self, $cb, $eol) = @_;
578 my $cb = pop;
579 my $eol = @_ ? shift : qr|(\015?\012)|;
580 my $pos;
581 1338
1339 if (@_ < 3) {
1340 # this is more than twice as fast as the generic code below
1341 sub {
1342 $_[0]{rbuf} =~ s/^([^\015\012]*)(\015?\012)// or return;
1343
1344 $cb->($_[0], $1, $2);
1345 1
1346 }
1347 } else {
582 $eol = quotemeta $eol unless ref $eol; 1348 $eol = quotemeta $eol unless ref $eol;
583 $eol = qr|^(.*?)($eol)|s; 1349 $eol = qr|^(.*?)($eol)|s;
1350
1351 sub {
1352 $_[0]{rbuf} =~ s/$eol// or return;
1353
1354 $cb->($_[0], $1, $2);
1355 1
1356 }
1357 }
1358};
1359
1360=item regex => $accept[, $reject[, $skip], $cb->($handle, $data)
1361
1362Makes a regex match against the regex object C<$accept> and returns
1363everything up to and including the match.
1364
1365Example: read a single line terminated by '\n'.
1366
1367 $handle->push_read (regex => qr<\n>, sub { ... });
1368
1369If C<$reject> is given and not undef, then it determines when the data is
1370to be rejected: it is matched against the data when the C<$accept> regex
1371does not match and generates an C<EBADMSG> error when it matches. This is
1372useful to quickly reject wrong data (to avoid waiting for a timeout or a
1373receive buffer overflow).
1374
1375Example: expect a single decimal number followed by whitespace, reject
1376anything else (not the use of an anchor).
1377
1378 $handle->push_read (regex => qr<^[0-9]+\s>, qr<[^0-9]>, sub { ... });
1379
1380If C<$skip> is given and not C<undef>, then it will be matched against
1381the receive buffer when neither C<$accept> nor C<$reject> match,
1382and everything preceding and including the match will be accepted
1383unconditionally. This is useful to skip large amounts of data that you
1384know cannot be matched, so that the C<$accept> or C<$reject> regex do not
1385have to start matching from the beginning. This is purely an optimisation
1386and is usually worth only when you expect more than a few kilobytes.
1387
1388Example: expect a http header, which ends at C<\015\012\015\012>. Since we
1389expect the header to be very large (it isn't in practise, but...), we use
1390a skip regex to skip initial portions. The skip regex is tricky in that
1391it only accepts something not ending in either \015 or \012, as these are
1392required for the accept regex.
1393
1394 $handle->push_read (regex =>
1395 qr<\015\012\015\012>,
1396 undef, # no reject
1397 qr<^.*[^\015\012]>,
1398 sub { ... });
1399
1400=cut
1401
1402register_read_type regex => sub {
1403 my ($self, $cb, $accept, $reject, $skip) = @_;
1404
1405 my $data;
1406 my $rbuf = \$self->{rbuf};
584 1407
585 sub { 1408 sub {
586 $_[0]{rbuf} =~ s/$eol// or return; 1409 # accept
1410 if ($$rbuf =~ $accept) {
1411 $data .= substr $$rbuf, 0, $+[0], "";
1412 $cb->($self, $data);
1413 return 1;
1414 }
1415
1416 # reject
1417 if ($reject && $$rbuf =~ $reject) {
1418 $self->_error (Errno::EBADMSG);
1419 }
587 1420
588 $cb->($_[0], $1, $2); 1421 # skip
1422 if ($skip && $$rbuf =~ $skip) {
1423 $data .= substr $$rbuf, 0, $+[0], "";
1424 }
1425
1426 ()
1427 }
1428};
1429
1430=item netstring => $cb->($handle, $string)
1431
1432A netstring (http://cr.yp.to/proto/netstrings.txt, this is not an endorsement).
1433
1434Throws an error with C<$!> set to EBADMSG on format violations.
1435
1436=cut
1437
1438register_read_type netstring => sub {
1439 my ($self, $cb) = @_;
1440
1441 sub {
1442 unless ($_[0]{rbuf} =~ s/^(0|[1-9][0-9]*)://) {
1443 if ($_[0]{rbuf} =~ /[^0-9]/) {
1444 $self->_error (Errno::EBADMSG);
1445 }
1446 return;
1447 }
1448
1449 my $len = $1;
1450
1451 $self->unshift_read (chunk => $len, sub {
1452 my $string = $_[1];
1453 $_[0]->unshift_read (chunk => 1, sub {
1454 if ($_[1] eq ",") {
1455 $cb->($_[0], $string);
1456 } else {
1457 $self->_error (Errno::EBADMSG);
1458 }
1459 });
1460 });
1461
589 1 1462 1
590 } 1463 }
591} 1464};
592 1465
593sub push_read_line { 1466=item packstring => $format, $cb->($handle, $string)
594 $_[0]->push_read (&_read_line);
595}
596 1467
597sub unshift_read_line { 1468An octet string prefixed with an encoded length. The encoding C<$format>
598 $_[0]->unshift_read (&_read_line); 1469uses the same format as a Perl C<pack> format, but must specify a single
599} 1470integer only (only one of C<cCsSlLqQiInNvVjJw> is allowed, plus an
1471optional C<!>, C<< < >> or C<< > >> modifier).
1472
1473For example, DNS over TCP uses a prefix of C<n> (2 octet network order),
1474EPP uses a prefix of C<N> (4 octtes).
1475
1476Example: read a block of data prefixed by its length in BER-encoded
1477format (very efficient).
1478
1479 $handle->push_read (packstring => "w", sub {
1480 my ($handle, $data) = @_;
1481 });
1482
1483=cut
1484
1485register_read_type packstring => sub {
1486 my ($self, $cb, $format) = @_;
1487
1488 sub {
1489 # when we can use 5.10 we can use ".", but for 5.8 we use the re-pack method
1490 defined (my $len = eval { unpack $format, $_[0]{rbuf} })
1491 or return;
1492
1493 $format = length pack $format, $len;
1494
1495 # bypass unshift if we already have the remaining chunk
1496 if ($format + $len <= length $_[0]{rbuf}) {
1497 my $data = substr $_[0]{rbuf}, $format, $len;
1498 substr $_[0]{rbuf}, 0, $format + $len, "";
1499 $cb->($_[0], $data);
1500 } else {
1501 # remove prefix
1502 substr $_[0]{rbuf}, 0, $format, "";
1503
1504 # read remaining chunk
1505 $_[0]->unshift_read (chunk => $len, $cb);
1506 }
1507
1508 1
1509 }
1510};
1511
1512=item json => $cb->($handle, $hash_or_arrayref)
1513
1514Reads a JSON object or array, decodes it and passes it to the
1515callback. When a parse error occurs, an C<EBADMSG> error will be raised.
1516
1517If a C<json> object was passed to the constructor, then that will be used
1518for the final decode, otherwise it will create a JSON coder expecting UTF-8.
1519
1520This read type uses the incremental parser available with JSON version
15212.09 (and JSON::XS version 2.2) and above. You have to provide a
1522dependency on your own: this module will load the JSON module, but
1523AnyEvent does not depend on it itself.
1524
1525Since JSON texts are fully self-delimiting, the C<json> read and write
1526types are an ideal simple RPC protocol: just exchange JSON datagrams. See
1527the C<json> write type description, above, for an actual example.
1528
1529=cut
1530
1531register_read_type json => sub {
1532 my ($self, $cb) = @_;
1533
1534 my $json = $self->{json} ||= json_coder;
1535
1536 my $data;
1537 my $rbuf = \$self->{rbuf};
1538
1539 sub {
1540 my $ref = eval { $json->incr_parse ($self->{rbuf}) };
1541
1542 if ($ref) {
1543 $self->{rbuf} = $json->incr_text;
1544 $json->incr_text = "";
1545 $cb->($self, $ref);
1546
1547 1
1548 } elsif ($@) {
1549 # error case
1550 $json->incr_skip;
1551
1552 $self->{rbuf} = $json->incr_text;
1553 $json->incr_text = "";
1554
1555 $self->_error (Errno::EBADMSG);
1556
1557 ()
1558 } else {
1559 $self->{rbuf} = "";
1560
1561 ()
1562 }
1563 }
1564};
1565
1566=item storable => $cb->($handle, $ref)
1567
1568Deserialises a L<Storable> frozen representation as written by the
1569C<storable> write type (BER-encoded length prefix followed by nfreeze'd
1570data).
1571
1572Raises C<EBADMSG> error if the data could not be decoded.
1573
1574=cut
1575
1576register_read_type storable => sub {
1577 my ($self, $cb) = @_;
1578
1579 require Storable;
1580
1581 sub {
1582 # when we can use 5.10 we can use ".", but for 5.8 we use the re-pack method
1583 defined (my $len = eval { unpack "w", $_[0]{rbuf} })
1584 or return;
1585
1586 my $format = length pack "w", $len;
1587
1588 # bypass unshift if we already have the remaining chunk
1589 if ($format + $len <= length $_[0]{rbuf}) {
1590 my $data = substr $_[0]{rbuf}, $format, $len;
1591 substr $_[0]{rbuf}, 0, $format + $len, "";
1592 $cb->($_[0], Storable::thaw ($data));
1593 } else {
1594 # remove prefix
1595 substr $_[0]{rbuf}, 0, $format, "";
1596
1597 # read remaining chunk
1598 $_[0]->unshift_read (chunk => $len, sub {
1599 if (my $ref = eval { Storable::thaw ($_[1]) }) {
1600 $cb->($_[0], $ref);
1601 } else {
1602 $self->_error (Errno::EBADMSG);
1603 }
1604 });
1605 }
1606
1607 1
1608 }
1609};
1610
1611=back
1612
1613=item AnyEvent::Handle::register_read_type type => $coderef->($handle, $cb, @args)
1614
1615This function (not method) lets you add your own types to C<push_read>.
1616
1617Whenever the given C<type> is used, C<push_read> will invoke the code
1618reference with the handle object, the callback and the remaining
1619arguments.
1620
1621The code reference is supposed to return a callback (usually a closure)
1622that works as a plain read callback (see C<< ->push_read ($cb) >>).
1623
1624It should invoke the passed callback when it is done reading (remember to
1625pass C<$handle> as first argument as all other callbacks do that).
1626
1627Note that this is a function, and all types registered this way will be
1628global, so try to use unique names.
1629
1630For examples, see the source of this module (F<perldoc -m AnyEvent::Handle>,
1631search for C<register_read_type>)).
600 1632
601=item $handle->stop_read 1633=item $handle->stop_read
602 1634
603=item $handle->start_read 1635=item $handle->start_read
604 1636
605In rare cases you actually do not want to read anything from the 1637In rare cases you actually do not want to read anything from the
606socket. In this case you can call C<stop_read>. Neither C<on_read> no 1638socket. In this case you can call C<stop_read>. Neither C<on_read> nor
607any queued callbacks will be executed then. To start reading again, call 1639any queued callbacks will be executed then. To start reading again, call
608C<start_read>. 1640C<start_read>.
609 1641
1642Note that AnyEvent::Handle will automatically C<start_read> for you when
1643you change the C<on_read> callback or push/unshift a read callback, and it
1644will automatically C<stop_read> for you when neither C<on_read> is set nor
1645there are any read requests in the queue.
1646
1647These methods will have no effect when in TLS mode (as TLS doesn't support
1648half-duplex connections).
1649
610=cut 1650=cut
611 1651
612sub stop_read { 1652sub stop_read {
613 my ($self) = @_; 1653 my ($self) = @_;
614 1654
615 delete $self->{rw}; 1655 delete $self->{_rw} unless $self->{tls};
616} 1656}
617 1657
618sub start_read { 1658sub start_read {
619 my ($self) = @_; 1659 my ($self) = @_;
620 1660
621 unless ($self->{rw} || $self->{eof}) { 1661 unless ($self->{_rw} || $self->{_eof}) {
622 Scalar::Util::weaken $self; 1662 Scalar::Util::weaken $self;
623 1663
624 $self->{rw} = AnyEvent->io (fh => $self->{fh}, poll => "r", cb => sub { 1664 $self->{_rw} = AE::io $self->{fh}, 0, sub {
625 my $rbuf = $self->{filter_r} ? \my $buf : \$self->{rbuf}; 1665 my $rbuf = \($self->{tls} ? my $buf : $self->{rbuf});
626 my $len = sysread $self->{fh}, $$rbuf, $self->{read_size} || 8192, length $$rbuf; 1666 my $len = sysread $self->{fh}, $$rbuf, $self->{read_size} || 8192, length $$rbuf;
627 1667
628 if ($len > 0) { 1668 if ($len > 0) {
629 $self->{filter_r} 1669 $self->{_activity} = $self->{_ractivity} = AE::now;
630 ? $self->{filter_r}->($self, $rbuf) 1670
1671 if ($self->{tls}) {
1672 Net::SSLeay::BIO_write ($self->{_rbio}, $$rbuf);
1673
1674 &_dotls ($self);
1675 } else {
631 : $self->_drain_rbuf; 1676 $self->_drain_rbuf;
1677 }
632 1678
633 } elsif (defined $len) { 1679 } elsif (defined $len) {
634 delete $self->{rw}; 1680 delete $self->{_rw};
635 $self->{eof} = 1; 1681 $self->{_eof} = 1;
636 $self->_drain_rbuf; 1682 $self->_drain_rbuf;
637 1683
638 } elsif ($! != EAGAIN && $! != EINTR) { 1684 } elsif ($! != EAGAIN && $! != EINTR && $! != WSAEWOULDBLOCK) {
639 return $self->error; 1685 return $self->_error ($!, 1);
640 } 1686 }
641 }); 1687 };
642 } 1688 }
643} 1689}
644 1690
1691our $ERROR_SYSCALL;
1692our $ERROR_WANT_READ;
1693
1694sub _tls_error {
1695 my ($self, $err) = @_;
1696
1697 return $self->_error ($!, 1)
1698 if $err == Net::SSLeay::ERROR_SYSCALL ();
1699
1700 my $err =Net::SSLeay::ERR_error_string (Net::SSLeay::ERR_get_error ());
1701
1702 # reduce error string to look less scary
1703 $err =~ s/^error:[0-9a-fA-F]{8}:[^:]+:([^:]+):/\L$1: /;
1704
1705 if ($self->{_on_starttls}) {
1706 (delete $self->{_on_starttls})->($self, undef, $err);
1707 &_freetls;
1708 } else {
1709 &_freetls;
1710 $self->_error (Errno::EPROTO, 1, $err);
1711 }
1712}
1713
1714# poll the write BIO and send the data if applicable
1715# also decode read data if possible
1716# this is basiclaly our TLS state machine
1717# more efficient implementations are possible with openssl,
1718# but not with the buggy and incomplete Net::SSLeay.
645sub _dotls { 1719sub _dotls {
646 my ($self) = @_; 1720 my ($self) = @_;
647 1721
1722 my $tmp;
1723
648 if (length $self->{tls_wbuf}) { 1724 if (length $self->{_tls_wbuf}) {
649 while ((my $len = Net::SSLeay::write ($self->{tls}, $self->{tls_wbuf})) > 0) { 1725 while (($tmp = Net::SSLeay::write ($self->{tls}, $self->{_tls_wbuf})) > 0) {
650 substr $self->{tls_wbuf}, 0, $len, ""; 1726 substr $self->{_tls_wbuf}, 0, $tmp, "";
651 } 1727 }
652 }
653 1728
1729 $tmp = Net::SSLeay::get_error ($self->{tls}, $tmp);
1730 return $self->_tls_error ($tmp)
1731 if $tmp != $ERROR_WANT_READ
1732 && ($tmp != $ERROR_SYSCALL || $!);
1733 }
1734
654 if (defined (my $buf = Net::SSLeay::BIO_read ($self->{tls_wbio}))) { 1735 while (defined ($tmp = Net::SSLeay::read ($self->{tls}))) {
1736 unless (length $tmp) {
1737 $self->{_on_starttls}
1738 and (delete $self->{_on_starttls})->($self, undef, "EOF during handshake"); # ???
1739 &_freetls;
1740
1741 if ($self->{on_stoptls}) {
1742 $self->{on_stoptls}($self);
1743 return;
1744 } else {
1745 # let's treat SSL-eof as we treat normal EOF
1746 delete $self->{_rw};
1747 $self->{_eof} = 1;
1748 }
1749 }
1750
1751 $self->{_tls_rbuf} .= $tmp;
1752 $self->_drain_rbuf;
1753 $self->{tls} or return; # tls session might have gone away in callback
1754 }
1755
1756 $tmp = Net::SSLeay::get_error ($self->{tls}, -1);
1757 return $self->_tls_error ($tmp)
1758 if $tmp != $ERROR_WANT_READ
1759 && ($tmp != $ERROR_SYSCALL || $!);
1760
1761 while (length ($tmp = Net::SSLeay::BIO_read ($self->{_wbio}))) {
655 $self->{wbuf} .= $buf; 1762 $self->{wbuf} .= $tmp;
656 $self->_drain_wbuf; 1763 $self->_drain_wbuf;
657 } 1764 }
658 1765
659 while (defined (my $buf = Net::SSLeay::read ($self->{tls}))) { 1766 $self->{_on_starttls}
660 $self->{rbuf} .= $buf; 1767 and Net::SSLeay::state ($self->{tls}) == Net::SSLeay::ST_OK ()
661 $self->_drain_rbuf; 1768 and (delete $self->{_on_starttls})->($self, 1, "TLS/SSL connection established");
662 } 1769}
663 1770
664 my $err = Net::SSLeay::get_error ($self->{tls}, -1); 1771=item $handle->starttls ($tls[, $tls_ctx])
665 1772
666 if ($err!= Net::SSLeay::ERROR_WANT_READ ()) { 1773Instead of starting TLS negotiation immediately when the AnyEvent::Handle
667 if ($err == Net::SSLeay::ERROR_SYSCALL ()) { 1774object is created, you can also do that at a later time by calling
668 $self->error; 1775C<starttls>.
669 } elsif ($err == Net::SSLeay::ERROR_SSL ()) { 1776
670 $! = &Errno::EIO; 1777Starting TLS is currently an asynchronous operation - when you push some
671 $self->error; 1778write data and then call C<< ->starttls >> then TLS negotiation will start
1779immediately, after which the queued write data is then sent.
1780
1781The first argument is the same as the C<tls> constructor argument (either
1782C<"connect">, C<"accept"> or an existing Net::SSLeay object).
1783
1784The second argument is the optional C<AnyEvent::TLS> object that is used
1785when AnyEvent::Handle has to create its own TLS connection object, or
1786a hash reference with C<< key => value >> pairs that will be used to
1787construct a new context.
1788
1789The TLS connection object will end up in C<< $handle->{tls} >>, the TLS
1790context in C<< $handle->{tls_ctx} >> after this call and can be used or
1791changed to your liking. Note that the handshake might have already started
1792when this function returns.
1793
1794Due to bugs in OpenSSL, it might or might not be possible to do multiple
1795handshakes on the same stream. Best do not attempt to use the stream after
1796stopping TLS.
1797
1798=cut
1799
1800our %TLS_CACHE; #TODO not yet documented, should we?
1801
1802sub starttls {
1803 my ($self, $tls, $ctx) = @_;
1804
1805 Carp::croak "It is an error to call starttls on an AnyEvent::Handle object while TLS is already active, caught"
1806 if $self->{tls};
1807
1808 $self->{tls} = $tls;
1809 $self->{tls_ctx} = $ctx if @_ > 2;
1810
1811 return unless $self->{fh};
1812
1813 require Net::SSLeay;
1814
1815 $ERROR_SYSCALL = Net::SSLeay::ERROR_SYSCALL ();
1816 $ERROR_WANT_READ = Net::SSLeay::ERROR_WANT_READ ();
1817
1818 $tls = delete $self->{tls};
1819 $ctx = $self->{tls_ctx};
1820
1821 local $Carp::CarpLevel = 1; # skip ourselves when creating a new context or session
1822
1823 if ("HASH" eq ref $ctx) {
1824 require AnyEvent::TLS;
1825
1826 if ($ctx->{cache}) {
1827 my $key = $ctx+0;
1828 $ctx = $TLS_CACHE{$key} ||= new AnyEvent::TLS %$ctx;
1829 } else {
1830 $ctx = new AnyEvent::TLS %$ctx;
672 } 1831 }
673 1832 }
674 # all others are fine for our purposes
675 } 1833
676} 1834 $self->{tls_ctx} = $ctx || TLS_CTX ();
677 1835 $self->{tls} = $tls = $self->{tls_ctx}->_get_session ($tls, $self, $self->{peername});
678# TODO: maybe document...
679sub starttls {
680 my ($self, $ssl, $ctx) = @_;
681
682 if ($ssl eq "accept") {
683 $ssl = Net::SSLeay::new ($ctx || TLS_CTX ());
684 Net::SSLeay::set_accept_state ($ssl);
685 } elsif ($ssl eq "connect") {
686 $ssl = Net::SSLeay::new ($ctx || TLS_CTX ());
687 Net::SSLeay::set_connect_state ($ssl);
688 }
689
690 $self->{tls} = $ssl;
691 1836
692 # basically, this is deep magic (because SSL_read should have the same issues) 1837 # basically, this is deep magic (because SSL_read should have the same issues)
693 # but the openssl maintainers basically said: "trust us, it just works". 1838 # but the openssl maintainers basically said: "trust us, it just works".
694 # (unfortunately, we have to hardcode constants because the abysmally misdesigned 1839 # (unfortunately, we have to hardcode constants because the abysmally misdesigned
695 # and mismaintained ssleay-module doesn't even offer them). 1840 # and mismaintained ssleay-module doesn't even offer them).
1841 # http://www.mail-archive.com/openssl-dev@openssl.org/msg22420.html
1842 #
1843 # in short: this is a mess.
1844 #
1845 # note that we do not try to keep the length constant between writes as we are required to do.
1846 # we assume that most (but not all) of this insanity only applies to non-blocking cases,
1847 # and we drive openssl fully in blocking mode here. Or maybe we don't - openssl seems to
1848 # have identity issues in that area.
696 Net::SSLeay::CTX_set_mode ($self->{tls}, 1849# Net::SSLeay::CTX_set_mode ($ssl,
697 (eval { Net::SSLeay::MODE_ENABLE_PARTIAL_WRITE () } || 1) 1850# (eval { local $SIG{__DIE__}; Net::SSLeay::MODE_ENABLE_PARTIAL_WRITE () } || 1)
698 | (eval { Net::SSLeay::MODE_ACCEPT_MOVING_WRITE_BUFFER () } || 2)); 1851# | (eval { local $SIG{__DIE__}; Net::SSLeay::MODE_ACCEPT_MOVING_WRITE_BUFFER () } || 2));
1852 Net::SSLeay::CTX_set_mode ($tls, 1|2);
699 1853
700 $self->{tls_rbio} = Net::SSLeay::BIO_new (Net::SSLeay::BIO_s_mem ()); 1854 $self->{_rbio} = Net::SSLeay::BIO_new (Net::SSLeay::BIO_s_mem ());
701 $self->{tls_wbio} = Net::SSLeay::BIO_new (Net::SSLeay::BIO_s_mem ()); 1855 $self->{_wbio} = Net::SSLeay::BIO_new (Net::SSLeay::BIO_s_mem ());
702 1856
1857 Net::SSLeay::BIO_write ($self->{_rbio}, delete $self->{rbuf});
1858
703 Net::SSLeay::set_bio ($ssl, $self->{tls_rbio}, $self->{tls_wbio}); 1859 Net::SSLeay::set_bio ($tls, $self->{_rbio}, $self->{_wbio});
704 1860
705 $self->{filter_w} = sub { 1861 $self->{_on_starttls} = sub { $_[0]{on_starttls}(@_) }
706 $_[0]{tls_wbuf} .= ${$_[1]}; 1862 if $self->{on_starttls};
1863
1864 &_dotls; # need to trigger the initial handshake
1865 $self->start_read; # make sure we actually do read
1866}
1867
1868=item $handle->stoptls
1869
1870Shuts down the SSL connection - this makes a proper EOF handshake by
1871sending a close notify to the other side, but since OpenSSL doesn't
1872support non-blocking shut downs, it is not guarenteed that you can re-use
1873the stream afterwards.
1874
1875=cut
1876
1877sub stoptls {
1878 my ($self) = @_;
1879
1880 if ($self->{tls}) {
1881 Net::SSLeay::shutdown ($self->{tls});
1882
707 &_dotls; 1883 &_dotls;
1884
1885# # we don't give a shit. no, we do, but we can't. no...#d#
1886# # we, we... have to use openssl :/#d#
1887# &_freetls;#d#
1888 }
1889}
1890
1891sub _freetls {
1892 my ($self) = @_;
1893
1894 return unless $self->{tls};
1895
1896 $self->{tls_ctx}->_put_session (delete $self->{tls})
1897 if $self->{tls} > 0;
1898
1899 delete @$self{qw(_rbio _wbio _tls_wbuf _on_starttls)};
1900}
1901
1902sub DESTROY {
1903 my ($self) = @_;
1904
1905 &_freetls;
1906
1907 my $linger = exists $self->{linger} ? $self->{linger} : 3600;
1908
1909 if ($linger && length $self->{wbuf} && $self->{fh}) {
1910 my $fh = delete $self->{fh};
1911 my $wbuf = delete $self->{wbuf};
1912
1913 my @linger;
1914
1915 push @linger, AE::io $fh, 1, sub {
1916 my $len = syswrite $fh, $wbuf, length $wbuf;
1917
1918 if ($len > 0) {
1919 substr $wbuf, 0, $len, "";
1920 } else {
1921 @linger = (); # end
1922 }
1923 };
1924 push @linger, AE::timer $linger, 0, sub {
1925 @linger = ();
1926 };
1927 }
1928}
1929
1930=item $handle->destroy
1931
1932Shuts down the handle object as much as possible - this call ensures that
1933no further callbacks will be invoked and as many resources as possible
1934will be freed. Any method you will call on the handle object after
1935destroying it in this way will be silently ignored (and it will return the
1936empty list).
1937
1938Normally, you can just "forget" any references to an AnyEvent::Handle
1939object and it will simply shut down. This works in fatal error and EOF
1940callbacks, as well as code outside. It does I<NOT> work in a read or write
1941callback, so when you want to destroy the AnyEvent::Handle object from
1942within such an callback. You I<MUST> call C<< ->destroy >> explicitly in
1943that case.
1944
1945Destroying the handle object in this way has the advantage that callbacks
1946will be removed as well, so if those are the only reference holders (as
1947is common), then one doesn't need to do anything special to break any
1948reference cycles.
1949
1950The handle might still linger in the background and write out remaining
1951data, as specified by the C<linger> option, however.
1952
1953=cut
1954
1955sub destroy {
1956 my ($self) = @_;
1957
1958 $self->DESTROY;
1959 %$self = ();
1960 bless $self, "AnyEvent::Handle::destroyed";
1961}
1962
1963sub AnyEvent::Handle::destroyed::AUTOLOAD {
1964 #nop
1965}
1966
1967=item AnyEvent::Handle::TLS_CTX
1968
1969This function creates and returns the AnyEvent::TLS object used by default
1970for TLS mode.
1971
1972The context is created by calling L<AnyEvent::TLS> without any arguments.
1973
1974=cut
1975
1976our $TLS_CTX;
1977
1978sub TLS_CTX() {
1979 $TLS_CTX ||= do {
1980 require AnyEvent::TLS;
1981
1982 new AnyEvent::TLS
1983 }
1984}
1985
1986=back
1987
1988
1989=head1 NONFREQUENTLY ASKED QUESTIONS
1990
1991=over 4
1992
1993=item I C<undef> the AnyEvent::Handle reference inside my callback and
1994still get further invocations!
1995
1996That's because AnyEvent::Handle keeps a reference to itself when handling
1997read or write callbacks.
1998
1999It is only safe to "forget" the reference inside EOF or error callbacks,
2000from within all other callbacks, you need to explicitly call the C<<
2001->destroy >> method.
2002
2003=item I get different callback invocations in TLS mode/Why can't I pause
2004reading?
2005
2006Unlike, say, TCP, TLS connections do not consist of two independent
2007communication channels, one for each direction. Or put differently. The
2008read and write directions are not independent of each other: you cannot
2009write data unless you are also prepared to read, and vice versa.
2010
2011This can mean than, in TLS mode, you might get C<on_error> or C<on_eof>
2012callback invocations when you are not expecting any read data - the reason
2013is that AnyEvent::Handle always reads in TLS mode.
2014
2015During the connection, you have to make sure that you always have a
2016non-empty read-queue, or an C<on_read> watcher. At the end of the
2017connection (or when you no longer want to use it) you can call the
2018C<destroy> method.
2019
2020=item How do I read data until the other side closes the connection?
2021
2022If you just want to read your data into a perl scalar, the easiest way
2023to achieve this is by setting an C<on_read> callback that does nothing,
2024clearing the C<on_eof> callback and in the C<on_error> callback, the data
2025will be in C<$_[0]{rbuf}>:
2026
2027 $handle->on_read (sub { });
2028 $handle->on_eof (undef);
2029 $handle->on_error (sub {
2030 my $data = delete $_[0]{rbuf};
2031 });
2032
2033The reason to use C<on_error> is that TCP connections, due to latencies
2034and packets loss, might get closed quite violently with an error, when in
2035fact, all data has been received.
2036
2037It is usually better to use acknowledgements when transferring data,
2038to make sure the other side hasn't just died and you got the data
2039intact. This is also one reason why so many internet protocols have an
2040explicit QUIT command.
2041
2042=item I don't want to destroy the handle too early - how do I wait until
2043all data has been written?
2044
2045After writing your last bits of data, set the C<on_drain> callback
2046and destroy the handle in there - with the default setting of
2047C<low_water_mark> this will be called precisely when all data has been
2048written to the socket:
2049
2050 $handle->push_write (...);
2051 $handle->on_drain (sub {
2052 warn "all data submitted to the kernel\n";
2053 undef $handle;
2054 });
2055
2056If you just want to queue some data and then signal EOF to the other side,
2057consider using C<< ->push_shutdown >> instead.
2058
2059=item I want to contact a TLS/SSL server, I don't care about security.
2060
2061If your TLS server is a pure TLS server (e.g. HTTPS) that only speaks TLS,
2062simply connect to it and then create the AnyEvent::Handle with the C<tls>
2063parameter:
2064
2065 tcp_connect $host, $port, sub {
2066 my ($fh) = @_;
2067
2068 my $handle = new AnyEvent::Handle
2069 fh => $fh,
2070 tls => "connect",
2071 on_error => sub { ... };
2072
2073 $handle->push_write (...);
708 }; 2074 };
709 $self->{filter_r} = sub {
710 Net::SSLeay::BIO_write ($_[0]{tls_rbio}, ${$_[1]});
711 &_dotls;
712 };
713}
714 2075
715sub DESTROY { 2076=item I want to contact a TLS/SSL server, I do care about security.
716 my $self = shift;
717 2077
718 Net::SSLeay::free (delete $self->{tls}) if $self->{tls}; 2078Then you should additionally enable certificate verification, including
719} 2079peername verification, if the protocol you use supports it (see
2080L<AnyEvent::TLS>, C<verify_peername>).
720 2081
721=item AnyEvent::Handle::TLS_CTX 2082E.g. for HTTPS:
722 2083
723This function creates and returns the Net::SSLeay::CTX object used by 2084 tcp_connect $host, $port, sub {
724default for TLS mode. 2085 my ($fh) = @_;
725 2086
726The context is created like this: 2087 my $handle = new AnyEvent::Handle
2088 fh => $fh,
2089 peername => $host,
2090 tls => "connect",
2091 tls_ctx => { verify => 1, verify_peername => "https" },
2092 ...
727 2093
728 Net::SSLeay::load_error_strings; 2094Note that you must specify the hostname you connected to (or whatever
729 Net::SSLeay::SSLeay_add_ssl_algorithms; 2095"peername" the protocol needs) as the C<peername> argument, otherwise no
730 Net::SSLeay::randomize; 2096peername verification will be done.
731 2097
732 my $CTX = Net::SSLeay::CTX_new; 2098The above will use the system-dependent default set of trusted CA
2099certificates. If you want to check against a specific CA, add the
2100C<ca_file> (or C<ca_cert>) arguments to C<tls_ctx>:
733 2101
734 Net::SSLeay::CTX_set_options $CTX, Net::SSLeay::OP_ALL 2102 tls_ctx => {
2103 verify => 1,
2104 verify_peername => "https",
2105 ca_file => "my-ca-cert.pem",
2106 },
735 2107
736=cut 2108=item I want to create a TLS/SSL server, how do I do that?
737 2109
738our $TLS_CTX; 2110Well, you first need to get a server certificate and key. You have
2111three options: a) ask a CA (buy one, use cacert.org etc.) b) create a
2112self-signed certificate (cheap. check the search engine of your choice,
2113there are many tutorials on the net) or c) make your own CA (tinyca2 is a
2114nice program for that purpose).
739 2115
740sub TLS_CTX() { 2116Then create a file with your private key (in PEM format, see
741 $TLS_CTX || do { 2117L<AnyEvent::TLS>), followed by the certificate (also in PEM format). The
742 require Net::SSLeay; 2118file should then look like this:
743 2119
744 Net::SSLeay::load_error_strings (); 2120 -----BEGIN RSA PRIVATE KEY-----
745 Net::SSLeay::SSLeay_add_ssl_algorithms (); 2121 ...header data
746 Net::SSLeay::randomize (); 2122 ... lots of base64'y-stuff
2123 -----END RSA PRIVATE KEY-----
747 2124
748 $TLS_CTX = Net::SSLeay::CTX_new (); 2125 -----BEGIN CERTIFICATE-----
2126 ... lots of base64'y-stuff
2127 -----END CERTIFICATE-----
749 2128
750 Net::SSLeay::CTX_set_options ($TLS_CTX, Net::SSLeay::OP_ALL ()); 2129The important bits are the "PRIVATE KEY" and "CERTIFICATE" parts. Then
2130specify this file as C<cert_file>:
751 2131
752 $TLS_CTX 2132 tcp_server undef, $port, sub {
753 } 2133 my ($fh) = @_;
754} 2134
2135 my $handle = new AnyEvent::Handle
2136 fh => $fh,
2137 tls => "accept",
2138 tls_ctx => { cert_file => "my-server-keycert.pem" },
2139 ...
2140
2141When you have intermediate CA certificates that your clients might not
2142know about, just append them to the C<cert_file>.
755 2143
756=back 2144=back
757 2145
2146
2147=head1 SUBCLASSING AnyEvent::Handle
2148
2149In many cases, you might want to subclass AnyEvent::Handle.
2150
2151To make this easier, a given version of AnyEvent::Handle uses these
2152conventions:
2153
2154=over 4
2155
2156=item * all constructor arguments become object members.
2157
2158At least initially, when you pass a C<tls>-argument to the constructor it
2159will end up in C<< $handle->{tls} >>. Those members might be changed or
2160mutated later on (for example C<tls> will hold the TLS connection object).
2161
2162=item * other object member names are prefixed with an C<_>.
2163
2164All object members not explicitly documented (internal use) are prefixed
2165with an underscore character, so the remaining non-C<_>-namespace is free
2166for use for subclasses.
2167
2168=item * all members not documented here and not prefixed with an underscore
2169are free to use in subclasses.
2170
2171Of course, new versions of AnyEvent::Handle may introduce more "public"
2172member variables, but thats just life, at least it is documented.
2173
2174=back
2175
758=head1 AUTHOR 2176=head1 AUTHOR
759 2177
760Robin Redeker C<< <elmex at ta-sa.org> >>, Marc Lehmann <schmorp@schmorp.de>. 2178Robin Redeker C<< <elmex at ta-sa.org> >>, Marc Lehmann <schmorp@schmorp.de>.
761 2179
762=cut 2180=cut

Diff Legend

Removed lines
+ Added lines
< Changed lines
> Changed lines