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 1.39 by root, Tue May 27 04:59:51 2008 UTC vs.
Revision 1.239 by root, Tue Dec 10 20:39:12 2013 UTC

1package AnyEvent::Handle;
2
3no warnings;
4use strict;
5
6use AnyEvent ();
7use AnyEvent::Util qw(WSAWOULDBLOCK);
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 streaming handles via AnyEvent
16
17=cut
18
19our $VERSION = '0.04';
20 4
21=head1 SYNOPSIS 5=head1 SYNOPSIS
22 6
23 use AnyEvent; 7 use AnyEvent;
24 use AnyEvent::Handle; 8 use AnyEvent::Handle;
25 9
26 my $cv = AnyEvent->condvar; 10 my $cv = AnyEvent->condvar;
27 11
28 my $handle = 12 my $hdl; $hdl = new AnyEvent::Handle
29 AnyEvent::Handle->new (
30 fh => \*STDIN, 13 fh => \*STDIN,
31 on_eof => sub { 14 on_error => sub {
32 $cv->broadcast; 15 my ($hdl, $fatal, $msg) = @_;
33 }, 16 AE::log error => $msg;
17 $hdl->destroy;
18 $cv->send;
34 ); 19 };
35 20
36 # send some request line 21 # send some request line
37 $handle->push_write ("getinfo\015\012"); 22 $hdl->push_write ("getinfo\015\012");
38 23
39 # read the response line 24 # read the response line
40 $handle->push_read (line => sub { 25 $hdl->push_read (line => sub {
41 my ($handle, $line) = @_; 26 my ($hdl, $line) = @_;
42 warn "read line <$line>\n"; 27 say "got line <$line>";
43 $cv->send; 28 $cv->send;
44 }); 29 });
45 30
46 $cv->recv; 31 $cv->recv;
47 32
48=head1 DESCRIPTION 33=head1 DESCRIPTION
49 34
50This 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
51filehandles. For utility functions for doing non-blocking connects and accepts 36stream-based filehandles (sockets, pipes, and other stream things).
52on sockets see L<AnyEvent::Util>.
53 37
38The L<AnyEvent::Intro> tutorial contains some well-documented
39AnyEvent::Handle examples.
40
54In the following, when the documentation refers to of "bytes" then this 41In the following, where the documentation refers to "bytes", it means
55means characters. As sysread and syswrite are used for all I/O, their 42characters. As sysread and syswrite are used for all I/O, their
56treatment of characters applies to this module as well. 43treatment of characters applies to this module as well.
44
45At the very minimum, you should specify C<fh> or C<connect>, and the
46C<on_error> callback.
57 47
58All callbacks will be invoked with the handle object as their first 48All callbacks will be invoked with the handle object as their first
59argument. 49argument.
60 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 = $AnyEvent::VERSION;
64
65sub _load_func($) {
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
61=head1 METHODS 82=head1 METHODS
62 83
63=over 4 84=over 4
64 85
65=item B<new (%args)> 86=item $handle = B<new> AnyEvent::Handle fh => $filehandle, key => value...
66 87
67The constructor supports these arguments (all as key => value pairs). 88The constructor supports these arguments (all as C<< key => value >> pairs).
68 89
69=over 4 90=over 4
70 91
71=item fh => $filehandle [MANDATORY] 92=item fh => $filehandle [C<fh> or C<connect> MANDATORY]
72 93
73The filehandle this L<AnyEvent::Handle> object will operate on. 94The filehandle this L<AnyEvent::Handle> object will operate on.
74
75NOTE: The filehandle will be set to non-blocking (using 95NOTE: The filehandle will be set to non-blocking mode (using
76AnyEvent::Util::fh_nonblocking). 96C<AnyEvent::Util::fh_nonblocking>) by the constructor and needs to stay in
97that mode.
77 98
78=item on_eof => $cb->($self) 99=item connect => [$host, $service] [C<fh> or C<connect> MANDATORY]
79 100
80Set the callback to be called on EOF. 101Try to connect to the specified host and service (port), using
102C<AnyEvent::Socket::tcp_connect>. The C<$host> additionally becomes the
103default C<peername>.
81 104
82While not mandatory, it is highly recommended to set an eof callback, 105You have to specify either this parameter, or C<fh>, above.
83otherwise you might end up with a closed socket while you are still
84waiting for data.
85 106
107It is possible to push requests on the read and write queues, and modify
108properties of the stream, even while AnyEvent::Handle is connecting.
109
110When this parameter is specified, then the C<on_prepare>,
111C<on_connect_error> and C<on_connect> callbacks will be called under the
112appropriate circumstances:
113
114=over 4
115
86=item on_error => $cb->($self) 116=item on_prepare => $cb->($handle)
87 117
118This (rarely used) callback is called before a new connection is
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
121file handle with parameters required for the actual connect (as opposed to
122settings that can be changed when the connection is already established).
123
124The return value of this callback should be the connect timeout value in
125seconds (or C<0>, or C<undef>, or the empty list, to indicate that the
126default timeout is to be used).
127
128=item on_connect => $cb->($handle, $host, $port, $retry->())
129
130This callback is called when a connection has been successfully established.
131
132The peer's numeric host and port (the socket peername) are passed as
133parameters, together with a retry callback. At the time it is called the
134read and write queues, EOF status, TLS status and similar properties of
135the handle will have been reset.
136
137It is not allowed to use the read or write queues while the handle object
138is connecting.
139
140If, for some reason, the handle is not acceptable, calling C<$retry> will
141continue with the next connection target (in case of multi-homed hosts or
142SRV records there can be multiple connection endpoints). The C<$retry>
143callback can be invoked after the connect callback returns, i.e. one can
144start a handshake and then decide to retry with the next host if the
145handshake fails.
146
147In most cases, you should ignore the C<$retry> parameter.
148
149=item on_connect_error => $cb->($handle, $message)
150
151This callback is called when the connection could not be
152established. C<$!> will contain the relevant error code, and C<$message> a
153message describing it (usually the same as C<"$!">).
154
155If this callback isn't specified, then C<on_error> will be called with a
156fatal error instead.
157
158=back
159
160=item on_error => $cb->($handle, $fatal, $message)
161
88This is the fatal error callback, that is called when, well, a fatal error 162This is the error callback, which is called when, well, some error
89occurs, such as not being able to resolve the hostname, failure to connect 163occured, such as not being able to resolve the hostname, failure to
90or a read error. 164connect, or a read error.
91 165
92The object will not be in a usable state when this callback has been 166Some errors are fatal (which is indicated by C<$fatal> being true). On
93called. 167fatal errors the handle object will be destroyed (by a call to C<< ->
168destroy >>) after invoking the error callback (which means you are free to
169examine the handle object). Examples of fatal errors are an EOF condition
170with active (but unsatisfiable) read watchers (C<EPIPE>) or I/O errors. In
171cases where the other side can close the connection at will, it is
172often easiest to not report C<EPIPE> errors in this callback.
94 173
174AnyEvent::Handle tries to find an appropriate error code for you to check
175against, but in some cases (TLS errors), this does not work well.
176
177If you report the error to the user, it is recommended to always output
178the C<$message> argument in human-readable error messages (you don't need
179to report C<"$!"> if you report C<$message>).
180
181If you want to react programmatically to the error, then looking at C<$!>
182and comparing it against some of the documented C<Errno> values is usually
183better than looking at the C<$message>.
184
185Non-fatal errors can be retried by returning, but it is recommended
186to simply ignore this parameter and instead abondon the handle object
187when this callback is invoked. Examples of non-fatal errors are timeouts
188C<ETIMEDOUT>) or badly-formatted data (C<EBADMSG>).
189
95On callback entrance, the value of C<$!> contains the operating system 190On entry to the callback, the value of C<$!> contains the operating
96error (or C<ENOSPC>, C<EPIPE> or C<EBADMSG>). 191system error code (or C<ENOSPC>, C<EPIPE>, C<ETIMEDOUT>, C<EBADMSG> or
97 192C<EPROTO>).
98The callback should throw an exception. If it returns, then
99AnyEvent::Handle will C<croak> for you.
100 193
101While not mandatory, it is I<highly> recommended to set this callback, as 194While not mandatory, it is I<highly> recommended to set this callback, as
102you will not be notified of errors otherwise. The default simply calls 195you will not be notified of errors otherwise. The default just calls
103die. 196C<croak>.
104 197
105=item on_read => $cb->($self) 198=item on_read => $cb->($handle)
106 199
107This sets the default read callback, which is called when data arrives 200This sets the default read callback, which is called when data arrives
108and no read request is in the queue. 201and no read request is in the queue (unlike read queue callbacks, this
202callback will only be called when at least one octet of data is in the
203read buffer).
109 204
110To access (and remove data from) the read buffer, use the C<< ->rbuf >> 205To access (and remove data from) the read buffer, use the C<< ->rbuf >>
111method or access the C<$self->{rbuf}> member directly. 206method or access the C<< $handle->{rbuf} >> member directly. Note that you
207must not enlarge or modify the read buffer, you can only remove data at
208the beginning from it.
112 209
210You can also call C<< ->push_read (...) >> or any other function that
211modifies the read queue. Or do both. Or ...
212
113When an EOF condition is detected then AnyEvent::Handle will first try to 213When an EOF condition is detected, AnyEvent::Handle will first try to
114feed all the remaining data to the queued callbacks and C<on_read> before 214feed all the remaining data to the queued callbacks and C<on_read> before
115calling the C<on_eof> callback. If no progress can be made, then a fatal 215calling the C<on_eof> callback. If no progress can be made, then a fatal
116error will be raised (with C<$!> set to C<EPIPE>). 216error will be raised (with C<$!> set to C<EPIPE>).
117 217
218Note that, unlike requests in the read queue, an C<on_read> callback
219doesn't mean you I<require> some data: if there is an EOF and there
220are outstanding read requests then an error will be flagged. With an
221C<on_read> callback, the C<on_eof> callback will be invoked.
222
223=item on_eof => $cb->($handle)
224
225Set the callback to be called when an end-of-file condition is detected,
226i.e. in the case of a socket, when the other side has closed the
227connection cleanly, and there are no outstanding read requests in the
228queue (if there are read requests, then an EOF counts as an unexpected
229connection close and will be flagged as an error).
230
231For sockets, this just means that the other side has stopped sending data,
232you can still try to write data, and, in fact, one can return from the EOF
233callback and continue writing data, as only the read part has been shut
234down.
235
236If an EOF condition has been detected but no C<on_eof> callback has been
237set, then a fatal error will be raised with C<$!> set to <0>.
238
118=item on_drain => $cb->() 239=item on_drain => $cb->($handle)
119 240
120This sets the callback that is called when the write buffer becomes empty 241This sets the callback that is called once when the write buffer becomes
121(or when the callback is set and the buffer is empty already). 242empty (and immediately when the handle object is created).
122 243
123To append to the write buffer, use the C<< ->push_write >> method. 244To append to the write buffer, use the C<< ->push_write >> method.
245
246This callback is useful when you don't want to put all of your write data
247into the queue at once, for example, when you want to write the contents
248of some file to the socket you might not want to read the whole file into
249memory and push it into the queue, but instead only read more data from
250the file when the write queue becomes empty.
251
252=item timeout => $fractional_seconds
253
254=item rtimeout => $fractional_seconds
255
256=item wtimeout => $fractional_seconds
257
258If non-zero, then these enables an "inactivity" timeout: whenever this
259many seconds pass without a successful read or write on the underlying
260file handle (or a call to C<timeout_reset>), the C<on_timeout> callback
261will be invoked (and if that one is missing, a non-fatal C<ETIMEDOUT>
262error will be raised).
263
264There are three variants of the timeouts that work independently of each
265other, for both read and write (triggered when nothing was read I<OR>
266written), just read (triggered when nothing was read), and just write:
267C<timeout>, C<rtimeout> and C<wtimeout>, with corresponding callbacks
268C<on_timeout>, C<on_rtimeout> and C<on_wtimeout>, and reset functions
269C<timeout_reset>, C<rtimeout_reset>, and C<wtimeout_reset>.
270
271Note that timeout processing is active even when you do not have any
272outstanding read or write requests: If you plan to keep the connection
273idle then you should disable the timeout temporarily or ignore the
274timeout in the corresponding C<on_timeout> callback, in which case
275AnyEvent::Handle will simply restart the timeout.
276
277Zero (the default) disables the corresponding timeout.
278
279=item on_timeout => $cb->($handle)
280
281=item on_rtimeout => $cb->($handle)
282
283=item on_wtimeout => $cb->($handle)
284
285Called whenever the inactivity timeout passes. If you return from this
286callback, then the timeout will be reset as if some activity had happened,
287so this condition is not fatal in any way.
124 288
125=item rbuf_max => <bytes> 289=item rbuf_max => <bytes>
126 290
127If defined, then a fatal error will be raised (with C<$!> set to C<ENOSPC>) 291If defined, then a fatal error will be raised (with C<$!> set to C<ENOSPC>)
128when the read buffer ever (strictly) exceeds this size. This is useful to 292when the read buffer ever (strictly) exceeds this size. This is useful to
129avoid denial-of-service attacks. 293avoid some forms of denial-of-service attacks.
130 294
131For example, a server accepting connections from untrusted sources should 295For example, a server accepting connections from untrusted sources should
132be configured to accept only so-and-so much data that it cannot act on 296be configured to accept only so-and-so much data that it cannot act on
133(for example, when expecting a line, an attacker could send an unlimited 297(for example, when expecting a line, an attacker could send an unlimited
134amount of data without a callback ever being called as long as the line 298amount of data without a callback ever being called as long as the line
135isn't finished). 299isn't finished).
136 300
301=item wbuf_max => <bytes>
302
303If defined, then a fatal error will be raised (with C<$!> set to C<ENOSPC>)
304when the write buffer ever (strictly) exceeds this size. This is useful to
305avoid some forms of denial-of-service attacks.
306
307Although the units of this parameter is bytes, this is the I<raw> number
308of bytes not yet accepted by the kernel. This can make a difference when
309you e.g. use TLS, as TLS typically makes your write data larger (but it
310can also make it smaller due to compression).
311
312As an example of when this limit is useful, take a chat server that sends
313chat messages to a client. If the client does not read those in a timely
314manner then the send buffer in the server would grow unbounded.
315
316=item autocork => <boolean>
317
318When disabled (the default), C<push_write> will try to immediately
319write the data to the handle if possible. This avoids having to register
320a write watcher and wait for the next event loop iteration, but can
321be inefficient if you write multiple small chunks (on the wire, this
322disadvantage is usually avoided by your kernel's nagle algorithm, see
323C<no_delay>, but this option can save costly syscalls).
324
325When enabled, writes will always be queued till the next event loop
326iteration. This is efficient when you do many small writes per iteration,
327but less efficient when you do a single write only per iteration (or when
328the write buffer often is full). It also increases write latency.
329
330=item no_delay => <boolean>
331
332When doing small writes on sockets, your operating system kernel might
333wait a bit for more data before actually sending it out. This is called
334the Nagle algorithm, and usually it is beneficial.
335
336In some situations you want as low a delay as possible, which can be
337accomplishd by setting this option to a true value.
338
339The default is your operating system's default behaviour (most likely
340enabled). This option explicitly enables or disables it, if possible.
341
342=item keepalive => <boolean>
343
344Enables (default disable) the SO_KEEPALIVE option on the stream socket:
345normally, TCP connections have no time-out once established, so TCP
346connections, once established, can stay alive forever even when the other
347side has long gone. TCP keepalives are a cheap way to take down long-lived
348TCP connections when the other side becomes unreachable. While the default
349is OS-dependent, TCP keepalives usually kick in after around two hours,
350and, if the other side doesn't reply, take down the TCP connection some 10
351to 15 minutes later.
352
353It is harmless to specify this option for file handles that do not support
354keepalives, and enabling it on connections that are potentially long-lived
355is usually a good idea.
356
357=item oobinline => <boolean>
358
359BSD majorly fucked up the implementation of TCP urgent data. The result
360is that almost no OS implements TCP according to the specs, and every OS
361implements it slightly differently.
362
363If you want to handle TCP urgent data, then setting this flag (the default
364is enabled) gives you the most portable way of getting urgent data, by
365putting it into the stream.
366
367Since BSD emulation of OOB data on top of TCP's urgent data can have
368security implications, AnyEvent::Handle sets this flag automatically
369unless explicitly specified. Note that setting this flag after
370establishing a connection I<may> be a bit too late (data loss could
371already have occured on BSD systems), but at least it will protect you
372from most attacks.
373
137=item read_size => <bytes> 374=item read_size => <bytes>
138 375
139The default read block size (the amount of bytes this module will try to read 376The initial read block size, the number of bytes this module will try
140on each [loop iteration). Default: C<4096>. 377to read during each loop iteration. Each handle object will consume
378at least this amount of memory for the read buffer as well, so when
379handling many connections watch out for memory requirements). See also
380C<max_read_size>. Default: C<2048>.
381
382=item max_read_size => <bytes>
383
384The maximum read buffer size used by the dynamic adjustment
385algorithm: Each time AnyEvent::Handle can read C<read_size> bytes in
386one go it will double C<read_size> up to the maximum given by this
387option. Default: C<131072> or C<read_size>, whichever is higher.
141 388
142=item low_water_mark => <bytes> 389=item low_water_mark => <bytes>
143 390
144Sets the amount of bytes (default: C<0>) that make up an "empty" write 391Sets the number of bytes (default: C<0>) that make up an "empty" write
145buffer: If the write reaches this size or gets even samller it is 392buffer: If the buffer reaches this size or gets even samller it is
146considered empty. 393considered empty.
147 394
395Sometimes it can be beneficial (for performance reasons) to add data to
396the write buffer before it is fully drained, but this is a rare case, as
397the operating system kernel usually buffers data as well, so the default
398is good in almost all cases.
399
400=item linger => <seconds>
401
402If this is non-zero (default: C<3600>), the destructor of the
403AnyEvent::Handle object will check whether there is still outstanding
404write data and will install a watcher that will write this data to the
405socket. No errors will be reported (this mostly matches how the operating
406system treats outstanding data at socket close time).
407
408This will not work for partial TLS data that could not be encoded
409yet. This data will be lost. Calling the C<stoptls> method in time might
410help.
411
412=item peername => $string
413
414A string used to identify the remote site - usually the DNS hostname
415(I<not> IDN!) used to create the connection, rarely the IP address.
416
417Apart from being useful in error messages, this string is also used in TLS
418peername verification (see C<verify_peername> in L<AnyEvent::TLS>). This
419verification will be skipped when C<peername> is not specified or is
420C<undef>.
421
148=item tls => "accept" | "connect" | Net::SSLeay::SSL object 422=item tls => "accept" | "connect" | Net::SSLeay::SSL object
149 423
150When this parameter is given, it enables TLS (SSL) mode, that means it 424When this parameter is given, it enables TLS (SSL) mode, that means
151will start making tls handshake and will transparently encrypt/decrypt 425AnyEvent will start a TLS handshake as soon as the connection has been
152data. 426established and will transparently encrypt/decrypt data afterwards.
427
428All TLS protocol errors will be signalled as C<EPROTO>, with an
429appropriate error message.
153 430
154TLS mode requires Net::SSLeay to be installed (it will be loaded 431TLS mode requires Net::SSLeay to be installed (it will be loaded
155automatically when you try to create a TLS handle). 432automatically when you try to create a TLS handle): this module doesn't
433have a dependency on that module, so if your module requires it, you have
434to add the dependency yourself. If Net::SSLeay cannot be loaded or is too
435old, you get an C<EPROTO> error.
156 436
157For the TLS server side, use C<accept>, and for the TLS client side of a 437Unlike TCP, TLS has a server and client side: for the TLS server side, use
158connection, use C<connect> mode. 438C<accept>, and for the TLS client side of a connection, use C<connect>
439mode.
159 440
160You can also provide your own TLS connection object, but you have 441You can also provide your own TLS connection object, but you have
161to make sure that you call either C<Net::SSLeay::set_connect_state> 442to make sure that you call either C<Net::SSLeay::set_connect_state>
162or C<Net::SSLeay::set_accept_state> on it before you pass it to 443or C<Net::SSLeay::set_accept_state> on it before you pass it to
163AnyEvent::Handle. 444AnyEvent::Handle. Also, this module will take ownership of this connection
445object.
164 446
447At some future point, AnyEvent::Handle might switch to another TLS
448implementation, then the option to use your own session object will go
449away.
450
451B<IMPORTANT:> since Net::SSLeay "objects" are really only integers,
452passing in the wrong integer will lead to certain crash. This most often
453happens when one uses a stylish C<< tls => 1 >> and is surprised about the
454segmentation fault.
455
165See the C<starttls> method if you need to start TLs negotiation later. 456Use the C<< ->starttls >> method if you need to start TLS negotiation later.
166 457
167=item tls_ctx => $ssl_ctx 458=item tls_ctx => $anyevent_tls
168 459
169Use the given Net::SSLeay::CTX object to create the new TLS connection 460Use the given C<AnyEvent::TLS> object to create the new TLS connection
170(unless a connection object was specified directly). If this parameter is 461(unless a connection object was specified directly). If this
171missing, then AnyEvent::Handle will use C<AnyEvent::Handle::TLS_CTX>. 462parameter is missing (or C<undef>), then AnyEvent::Handle will use
463C<AnyEvent::Handle::TLS_CTX>.
172 464
173=item filter_r => $cb 465Instead of an object, you can also specify a hash reference with C<< key
466=> value >> pairs. Those will be passed to L<AnyEvent::TLS> to create a
467new TLS context object.
174 468
175=item filter_w => $cb 469=item on_starttls => $cb->($handle, $success[, $error_message])
176 470
177These exist, but are undocumented at this time. 471This callback will be invoked when the TLS/SSL handshake has finished. If
472C<$success> is true, then the TLS handshake succeeded, otherwise it failed
473(C<on_stoptls> will not be called in this case).
474
475The session in C<< $handle->{tls} >> can still be examined in this
476callback, even when the handshake was not successful.
477
478TLS handshake failures will not cause C<on_error> to be invoked when this
479callback is in effect, instead, the error message will be passed to C<on_starttls>.
480
481Without this callback, handshake failures lead to C<on_error> being
482called as usual.
483
484Note that you cannot just call C<starttls> again in this callback. If you
485need to do that, start an zero-second timer instead whose callback can
486then call C<< ->starttls >> again.
487
488=item on_stoptls => $cb->($handle)
489
490When a SSLv3/TLS shutdown/close notify/EOF is detected and this callback is
491set, then it will be invoked after freeing the TLS session. If it is not,
492then a TLS shutdown condition will be treated like a normal EOF condition
493on the handle.
494
495The session in C<< $handle->{tls} >> can still be examined in this
496callback.
497
498This callback will only be called on TLS shutdowns, not when the
499underlying handle signals EOF.
500
501=item json => L<JSON> or L<JSON::XS> object
502
503This is the json coder object used by the C<json> read and write types.
504
505If you don't supply it, then AnyEvent::Handle will create and use a
506suitable one (on demand), which will write and expect UTF-8 encoded JSON
507texts.
508
509Note that you are responsible to depend on the L<JSON> module if you want
510to use this functionality, as AnyEvent does not have a dependency on it
511itself.
512
513=item cbor => L<CBOR::XS> object
514
515This is the cbor coder object used by the C<cbor> read and write types.
516
517If you don't supply it, then AnyEvent::Handle will create and use a
518suitable one (on demand), which will write CBOR without using extensions,
519if possible. texts.
520
521Note that you are responsible to depend on the L<CBOR::XS> module if you
522want to use this functionality, as AnyEvent does not have a dependency on
523it itself.
178 524
179=back 525=back
180 526
181=cut 527=cut
182 528
183sub new { 529sub new {
184 my $class = shift; 530 my $class = shift;
185
186 my $self = bless { @_ }, $class; 531 my $self = bless { @_ }, $class;
187 532
188 $self->{fh} or Carp::croak "mandatory argument fh is missing"; 533 if ($self->{fh}) {
534 $self->_start;
535 return unless $self->{fh}; # could be gone by now
536
537 } elsif ($self->{connect}) {
538 require AnyEvent::Socket;
539
540 $self->{peername} = $self->{connect}[0]
541 unless exists $self->{peername};
542
543 $self->{_skip_drain_rbuf} = 1;
544
545 {
546 Scalar::Util::weaken (my $self = $self);
547
548 $self->{_connect} =
549 AnyEvent::Socket::tcp_connect (
550 $self->{connect}[0],
551 $self->{connect}[1],
552 sub {
553 my ($fh, $host, $port, $retry) = @_;
554
555 delete $self->{_connect}; # no longer needed
556
557 if ($fh) {
558 $self->{fh} = $fh;
559
560 delete $self->{_skip_drain_rbuf};
561 $self->_start;
562
563 $self->{on_connect}
564 and $self->{on_connect}($self, $host, $port, sub {
565 delete @$self{qw(fh _tw _rtw _wtw _ww _rw _eof _queue rbuf _wbuf tls _tls_rbuf _tls_wbuf)};
566 $self->{_skip_drain_rbuf} = 1;
567 &$retry;
568 });
569
570 } else {
571 if ($self->{on_connect_error}) {
572 $self->{on_connect_error}($self, "$!");
573 $self->destroy if $self;
574 } else {
575 $self->_error ($!, 1);
576 }
577 }
578 },
579 sub {
580 local $self->{fh} = $_[0];
581
582 $self->{on_prepare}
583 ? $self->{on_prepare}->($self)
584 : ()
585 }
586 );
587 }
588
589 } else {
590 Carp::croak "AnyEvent::Handle: either an existing fh or the connect parameter must be specified";
591 }
592
593 $self
594}
595
596sub _start {
597 my ($self) = @_;
598
599 # too many clueless people try to use udp and similar sockets
600 # with AnyEvent::Handle, do them a favour.
601 my $type = getsockopt $self->{fh}, Socket::SOL_SOCKET (), Socket::SO_TYPE ();
602 Carp::croak "AnyEvent::Handle: only stream sockets supported, anything else will NOT work!"
603 if Socket::SOCK_STREAM () != (unpack "I", $type) && defined $type;
189 604
190 AnyEvent::Util::fh_nonblocking $self->{fh}, 1; 605 AnyEvent::Util::fh_nonblocking $self->{fh}, 1;
191 606
192 if ($self->{tls}) { 607 $self->{_activity} =
193 require Net::SSLeay; 608 $self->{_ractivity} =
609 $self->{_wactivity} = AE::now;
610
611 $self->{read_size} ||= 2048;
612 $self->{max_read_size} = $self->{read_size}
613 if $self->{read_size} > ($self->{max_read_size} || MAX_READ_SIZE);
614
615 $self->timeout (delete $self->{timeout} ) if $self->{timeout};
616 $self->rtimeout (delete $self->{rtimeout} ) if $self->{rtimeout};
617 $self->wtimeout (delete $self->{wtimeout} ) if $self->{wtimeout};
618
619 $self->no_delay (delete $self->{no_delay} ) if exists $self->{no_delay} && $self->{no_delay};
620 $self->keepalive (delete $self->{keepalive}) if exists $self->{keepalive} && $self->{keepalive};
621
622 $self->oobinline (exists $self->{oobinline} ? delete $self->{oobinline} : 1);
623
194 $self->starttls (delete $self->{tls}, delete $self->{tls_ctx}); 624 $self->starttls (delete $self->{tls}, delete $self->{tls_ctx})
195 } 625 if $self->{tls};
196 626
197 $self->on_eof (delete $self->{on_eof} ) if $self->{on_eof};
198 $self->on_error (delete $self->{on_error}) if $self->{on_error};
199 $self->on_drain (delete $self->{on_drain}) if $self->{on_drain}; 627 $self->on_drain (delete $self->{on_drain} ) if $self->{on_drain};
200 $self->on_read (delete $self->{on_read} ) if $self->{on_read};
201 628
202 $self->start_read; 629 $self->start_read
630 if $self->{on_read} || @{ $self->{_queue} };
203 631
204 $self 632 $self->_drain_wbuf;
205} 633}
206 634
207sub _shutdown {
208 my ($self) = @_;
209
210 delete $self->{_rw};
211 delete $self->{_ww};
212 delete $self->{fh};
213}
214
215sub error { 635sub _error {
216 my ($self) = @_; 636 my ($self, $errno, $fatal, $message) = @_;
217 637
218 { 638 $! = $errno;
219 local $!; 639 $message ||= "$!";
220 $self->_shutdown;
221 }
222 640
223 $self->{on_error}($self)
224 if $self->{on_error}; 641 if ($self->{on_error}) {
225 642 $self->{on_error}($self, $fatal, $message);
643 $self->destroy if $fatal;
644 } elsif ($self->{fh} || $self->{connect}) {
645 $self->destroy;
226 Carp::croak "AnyEvent::Handle uncaught fatal error: $!"; 646 Carp::croak "AnyEvent::Handle uncaught error: $message";
647 }
227} 648}
228 649
229=item $fh = $handle->fh 650=item $fh = $handle->fh
230 651
231This method returns the file handle of the L<AnyEvent::Handle> object. 652This method returns the file handle used to create the L<AnyEvent::Handle> object.
232 653
233=cut 654=cut
234 655
235sub fh { $_[0]{fh} } 656sub fh { $_[0]{fh} }
236 657
252 673
253sub on_eof { 674sub on_eof {
254 $_[0]{on_eof} = $_[1]; 675 $_[0]{on_eof} = $_[1];
255} 676}
256 677
678=item $handle->on_timeout ($cb)
679
680=item $handle->on_rtimeout ($cb)
681
682=item $handle->on_wtimeout ($cb)
683
684Replace the current C<on_timeout>, C<on_rtimeout> or C<on_wtimeout>
685callback, or disables the callback (but not the timeout) if C<$cb> =
686C<undef>. See the C<timeout> constructor argument and method.
687
688=cut
689
690# see below
691
692=item $handle->autocork ($boolean)
693
694Enables or disables the current autocork behaviour (see C<autocork>
695constructor argument). Changes will only take effect on the next write.
696
697=cut
698
699sub autocork {
700 $_[0]{autocork} = $_[1];
701}
702
703=item $handle->no_delay ($boolean)
704
705Enables or disables the C<no_delay> setting (see constructor argument of
706the same name for details).
707
708=cut
709
710sub no_delay {
711 $_[0]{no_delay} = $_[1];
712
713 setsockopt $_[0]{fh}, Socket::IPPROTO_TCP (), Socket::TCP_NODELAY (), int $_[1]
714 if $_[0]{fh};
715}
716
717=item $handle->keepalive ($boolean)
718
719Enables or disables the C<keepalive> setting (see constructor argument of
720the same name for details).
721
722=cut
723
724sub keepalive {
725 $_[0]{keepalive} = $_[1];
726
727 eval {
728 local $SIG{__DIE__};
729 setsockopt $_[0]{fh}, Socket::SOL_SOCKET (), Socket::SO_KEEPALIVE (), int $_[1]
730 if $_[0]{fh};
731 };
732}
733
734=item $handle->oobinline ($boolean)
735
736Enables or disables the C<oobinline> setting (see constructor argument of
737the same name for details).
738
739=cut
740
741sub oobinline {
742 $_[0]{oobinline} = $_[1];
743
744 eval {
745 local $SIG{__DIE__};
746 setsockopt $_[0]{fh}, Socket::SOL_SOCKET (), Socket::SO_OOBINLINE (), int $_[1]
747 if $_[0]{fh};
748 };
749}
750
751=item $handle->keepalive ($boolean)
752
753Enables or disables the C<keepalive> setting (see constructor argument of
754the same name for details).
755
756=cut
757
758sub keepalive {
759 $_[0]{keepalive} = $_[1];
760
761 eval {
762 local $SIG{__DIE__};
763 setsockopt $_[0]{fh}, Socket::SOL_SOCKET (), Socket::SO_KEEPALIVE (), int $_[1]
764 if $_[0]{fh};
765 };
766}
767
768=item $handle->on_starttls ($cb)
769
770Replace the current C<on_starttls> callback (see the C<on_starttls> constructor argument).
771
772=cut
773
774sub on_starttls {
775 $_[0]{on_starttls} = $_[1];
776}
777
778=item $handle->on_stoptls ($cb)
779
780Replace the current C<on_stoptls> callback (see the C<on_stoptls> constructor argument).
781
782=cut
783
784sub on_stoptls {
785 $_[0]{on_stoptls} = $_[1];
786}
787
788=item $handle->rbuf_max ($max_octets)
789
790Configures the C<rbuf_max> setting (C<undef> disables it).
791
792=item $handle->wbuf_max ($max_octets)
793
794Configures the C<wbuf_max> setting (C<undef> disables it).
795
796=cut
797
798sub rbuf_max {
799 $_[0]{rbuf_max} = $_[1];
800}
801
802sub wbuf_max {
803 $_[0]{wbuf_max} = $_[1];
804}
805
806#############################################################################
807
808=item $handle->timeout ($seconds)
809
810=item $handle->rtimeout ($seconds)
811
812=item $handle->wtimeout ($seconds)
813
814Configures (or disables) the inactivity timeout.
815
816The timeout will be checked instantly, so this method might destroy the
817handle before it returns.
818
819=item $handle->timeout_reset
820
821=item $handle->rtimeout_reset
822
823=item $handle->wtimeout_reset
824
825Reset the activity timeout, as if data was received or sent.
826
827These methods are cheap to call.
828
829=cut
830
831for my $dir ("", "r", "w") {
832 my $timeout = "${dir}timeout";
833 my $tw = "_${dir}tw";
834 my $on_timeout = "on_${dir}timeout";
835 my $activity = "_${dir}activity";
836 my $cb;
837
838 *$on_timeout = sub {
839 $_[0]{$on_timeout} = $_[1];
840 };
841
842 *$timeout = sub {
843 my ($self, $new_value) = @_;
844
845 $new_value >= 0
846 or Carp::croak "AnyEvent::Handle->$timeout called with negative timeout ($new_value), caught";
847
848 $self->{$timeout} = $new_value;
849 delete $self->{$tw}; &$cb;
850 };
851
852 *{"${dir}timeout_reset"} = sub {
853 $_[0]{$activity} = AE::now;
854 };
855
856 # main workhorse:
857 # reset the timeout watcher, as neccessary
858 # also check for time-outs
859 $cb = sub {
860 my ($self) = @_;
861
862 if ($self->{$timeout} && $self->{fh}) {
863 my $NOW = AE::now;
864
865 # when would the timeout trigger?
866 my $after = $self->{$activity} + $self->{$timeout} - $NOW;
867
868 # now or in the past already?
869 if ($after <= 0) {
870 $self->{$activity} = $NOW;
871
872 if ($self->{$on_timeout}) {
873 $self->{$on_timeout}($self);
874 } else {
875 $self->_error (Errno::ETIMEDOUT);
876 }
877
878 # callback could have changed timeout value, optimise
879 return unless $self->{$timeout};
880
881 # calculate new after
882 $after = $self->{$timeout};
883 }
884
885 Scalar::Util::weaken $self;
886 return unless $self; # ->error could have destroyed $self
887
888 $self->{$tw} ||= AE::timer $after, 0, sub {
889 delete $self->{$tw};
890 $cb->($self);
891 };
892 } else {
893 delete $self->{$tw};
894 }
895 }
896}
897
257############################################################################# 898#############################################################################
258 899
259=back 900=back
260 901
261=head2 WRITE QUEUE 902=head2 WRITE QUEUE
265 906
266The write queue is very simple: you can add data to its end, and 907The write queue is very simple: you can add data to its end, and
267AnyEvent::Handle will automatically try to get rid of it for you. 908AnyEvent::Handle will automatically try to get rid of it for you.
268 909
269When data could be written and the write buffer is shorter then the low 910When data could be written and the write buffer is shorter then the low
270water mark, the C<on_drain> callback will be invoked. 911water mark, the C<on_drain> callback will be invoked once.
271 912
272=over 4 913=over 4
273 914
274=item $handle->on_drain ($cb) 915=item $handle->on_drain ($cb)
275 916
276Sets the C<on_drain> callback or clears it (see the description of 917Sets the C<on_drain> callback or clears it (see the description of
277C<on_drain> in the constructor). 918C<on_drain> in the constructor).
278 919
920This method may invoke callbacks (and therefore the handle might be
921destroyed after it returns).
922
279=cut 923=cut
280 924
281sub on_drain { 925sub on_drain {
282 my ($self, $cb) = @_; 926 my ($self, $cb) = @_;
283 927
284 $self->{on_drain} = $cb; 928 $self->{on_drain} = $cb;
285 929
286 $cb->($self) 930 $cb->($self)
287 if $cb && $self->{low_water_mark} >= length $self->{wbuf}; 931 if $cb && $self->{low_water_mark} >= (length $self->{wbuf}) + (length $self->{_tls_wbuf});
288} 932}
289 933
290=item $handle->push_write ($data) 934=item $handle->push_write ($data)
291 935
292Queues the given scalar to be written. You can push as much data as you 936Queues the given scalar to be written. You can push as much data as
293want (only limited by the available memory), as C<AnyEvent::Handle> 937you want (only limited by the available memory and C<wbuf_max>), as
294buffers it independently of the kernel. 938C<AnyEvent::Handle> buffers it independently of the kernel.
939
940This method may invoke callbacks (and therefore the handle might be
941destroyed after it returns).
295 942
296=cut 943=cut
297 944
298sub _drain_wbuf { 945sub _drain_wbuf {
299 my ($self) = @_; 946 my ($self) = @_;
303 Scalar::Util::weaken $self; 950 Scalar::Util::weaken $self;
304 951
305 my $cb = sub { 952 my $cb = sub {
306 my $len = syswrite $self->{fh}, $self->{wbuf}; 953 my $len = syswrite $self->{fh}, $self->{wbuf};
307 954
308 if ($len >= 0) { 955 if (defined $len) {
309 substr $self->{wbuf}, 0, $len, ""; 956 substr $self->{wbuf}, 0, $len, "";
310 957
958 $self->{_activity} = $self->{_wactivity} = AE::now;
959
311 $self->{on_drain}($self) 960 $self->{on_drain}($self)
312 if $self->{low_water_mark} >= length $self->{wbuf} 961 if $self->{low_water_mark} >= (length $self->{wbuf}) + (length $self->{_tls_wbuf})
313 && $self->{on_drain}; 962 && $self->{on_drain};
314 963
315 delete $self->{_ww} unless length $self->{wbuf}; 964 delete $self->{_ww} unless length $self->{wbuf};
316 } elsif ($! != EAGAIN && $! != EINTR && $! != WSAWOULDBLOCK) { 965 } elsif ($! != EAGAIN && $! != EINTR && $! != WSAEWOULDBLOCK) {
317 $self->error; 966 $self->_error ($!, 1);
318 } 967 }
319 }; 968 };
320 969
321 # try to write data immediately 970 # try to write data immediately
322 $cb->(); 971 $cb->() unless $self->{autocork};
323 972
324 # if still data left in wbuf, we need to poll 973 # if still data left in wbuf, we need to poll
325 $self->{_ww} = AnyEvent->io (fh => $self->{fh}, poll => "w", cb => $cb) 974 $self->{_ww} = AE::io $self->{fh}, 1, $cb
326 if length $self->{wbuf}; 975 if length $self->{wbuf};
976
977 if (
978 defined $self->{wbuf_max}
979 && $self->{wbuf_max} < length $self->{wbuf}
980 ) {
981 $self->_error (Errno::ENOSPC, 1), return;
982 }
327 }; 983 };
328} 984}
329 985
330our %WH; 986our %WH;
331 987
988# deprecated
332sub register_write_type($$) { 989sub register_write_type($$) {
333 $WH{$_[0]} = $_[1]; 990 $WH{$_[0]} = $_[1];
334} 991}
335 992
336sub push_write { 993sub push_write {
337 my $self = shift; 994 my $self = shift;
338 995
339 if (@_ > 1) { 996 if (@_ > 1) {
340 my $type = shift; 997 my $type = shift;
341 998
999 @_ = ($WH{$type} ||= _load_func "$type\::anyevent_write_type"
342 @_ = ($WH{$type} or Carp::croak "unsupported type passed to AnyEvent::Handle::push_write") 1000 or Carp::croak "unsupported/unloadable type '$type' passed to AnyEvent::Handle::push_write")
343 ->($self, @_); 1001 ->($self, @_);
344 } 1002 }
345 1003
1004 # we downgrade here to avoid hard-to-track-down bugs,
1005 # and diagnose the problem earlier and better.
1006
346 if ($self->{filter_w}) { 1007 if ($self->{tls}) {
347 $self->{filter_w}->($self, \$_[0]); 1008 utf8::downgrade $self->{_tls_wbuf} .= $_[0];
1009 &_dotls ($self) if $self->{fh};
348 } else { 1010 } else {
349 $self->{wbuf} .= $_[0]; 1011 utf8::downgrade $self->{wbuf} .= $_[0];
350 $self->_drain_wbuf; 1012 $self->_drain_wbuf if $self->{fh};
351 } 1013 }
352} 1014}
353 1015
354=item $handle->push_write (type => @args) 1016=item $handle->push_write (type => @args)
355 1017
356=item $handle->unshift_write (type => @args)
357
358Instead of formatting your data yourself, you can also let this module do 1018Instead of formatting your data yourself, you can also let this module
359the job by specifying a type and type-specific arguments. 1019do the job by specifying a type and type-specific arguments. You
1020can also specify the (fully qualified) name of a package, in which
1021case AnyEvent tries to load the package and then expects to find the
1022C<anyevent_write_type> function inside (see "custom write types", below).
360 1023
361Predefined types are (if you have ideas for additional types, feel free to 1024Predefined types are (if you have ideas for additional types, feel free to
362drop by and tell us): 1025drop by and tell us):
363 1026
364=over 4 1027=over 4
366=item netstring => $string 1029=item netstring => $string
367 1030
368Formats the given value as netstring 1031Formats the given value as netstring
369(http://cr.yp.to/proto/netstrings.txt, this is not a recommendation to use them). 1032(http://cr.yp.to/proto/netstrings.txt, this is not a recommendation to use them).
370 1033
371=back
372
373=cut 1034=cut
374 1035
375register_write_type netstring => sub { 1036register_write_type netstring => sub {
376 my ($self, $string) = @_; 1037 my ($self, $string) = @_;
377 1038
378 sprintf "%d:%s,", (length $string), $string 1039 (length $string) . ":$string,"
379}; 1040};
380 1041
1042=item packstring => $format, $data
1043
1044An octet string prefixed with an encoded length. The encoding C<$format>
1045uses the same format as a Perl C<pack> format, but must specify a single
1046integer only (only one of C<cCsSlLqQiInNvVjJw> is allowed, plus an
1047optional C<!>, C<< < >> or C<< > >> modifier).
1048
1049=cut
1050
1051register_write_type packstring => sub {
1052 my ($self, $format, $string) = @_;
1053
1054 pack "$format/a*", $string
1055};
1056
381=item json => $array_or_hashref 1057=item json => $array_or_hashref
382 1058
383=item AnyEvent::Handle::register_write_type type => $coderef->($self, @args) 1059Encodes the given hash or array reference into a JSON object. Unless you
1060provide your own JSON object, this means it will be encoded to JSON text
1061in UTF-8.
384 1062
385This function (not method) lets you add your own types to C<push_write>. 1063JSON objects (and arrays) are self-delimiting, so you can write JSON at
1064one end of a handle and read them at the other end without using any
1065additional framing.
1066
1067The generated JSON text is guaranteed not to contain any newlines: While
1068this module doesn't need delimiters after or between JSON texts to be
1069able to read them, many other languages depend on that.
1070
1071A simple RPC protocol that interoperates easily with other languages is
1072to send JSON arrays (or objects, although arrays are usually the better
1073choice as they mimic how function argument passing works) and a newline
1074after each JSON text:
1075
1076 $handle->push_write (json => ["method", "arg1", "arg2"]); # whatever
1077 $handle->push_write ("\012");
1078
1079An AnyEvent::Handle receiver would simply use the C<json> read type and
1080rely on the fact that the newline will be skipped as leading whitespace:
1081
1082 $handle->push_read (json => sub { my $array = $_[1]; ... });
1083
1084Other languages could read single lines terminated by a newline and pass
1085this line into their JSON decoder of choice.
1086
1087=item cbor => $perl_scalar
1088
1089Encodes the given scalar into a CBOR value. Unless you provide your own
1090L<CBOR::XS> object, this means it will be encoded to a CBOR string not
1091using any extensions, if possible.
1092
1093CBOR values are self-delimiting, so you can write CBOR at one end of
1094a handle and read them at the other end without using any additional
1095framing.
1096
1097A simple nd very very fast RPC protocol that interoperates with
1098other languages is to send CBOR and receive CBOR values (arrays are
1099recommended):
1100
1101 $handle->push_write (cbor => ["method", "arg1", "arg2"]); # whatever
1102
1103An AnyEvent::Handle receiver would simply use the C<cbor> read type:
1104
1105 $handle->push_read (cbor => sub { my $array = $_[1]; ... });
1106
1107=cut
1108
1109sub json_coder() {
1110 eval { require JSON::XS; JSON::XS->new->utf8 }
1111 || do { require JSON; JSON->new->utf8 }
1112}
1113
1114register_write_type json => sub {
1115 my ($self, $ref) = @_;
1116
1117 ($self->{json} ||= json_coder)
1118 ->encode ($ref)
1119};
1120
1121sub cbor_coder() {
1122 require CBOR::XS;
1123 CBOR::XS->new
1124}
1125
1126register_write_type cbor => sub {
1127 my ($self, $scalar) = @_;
1128
1129 ($self->{cbor} ||= cbor_coder)
1130 ->encode ($scalar)
1131};
1132
1133=item storable => $reference
1134
1135Freezes the given reference using L<Storable> and writes it to the
1136handle. Uses the C<nfreeze> format.
1137
1138=cut
1139
1140register_write_type storable => sub {
1141 my ($self, $ref) = @_;
1142
1143 require Storable unless $Storable::VERSION;
1144
1145 pack "w/a*", Storable::nfreeze ($ref)
1146};
1147
1148=back
1149
1150=item $handle->push_shutdown
1151
1152Sometimes you know you want to close the socket after writing your data
1153before it was actually written. One way to do that is to replace your
1154C<on_drain> handler by a callback that shuts down the socket (and set
1155C<low_water_mark> to C<0>). This method is a shorthand for just that, and
1156replaces the C<on_drain> callback with:
1157
1158 sub { shutdown $_[0]{fh}, 1 }
1159
1160This simply shuts down the write side and signals an EOF condition to the
1161the peer.
1162
1163You can rely on the normal read queue and C<on_eof> handling
1164afterwards. This is the cleanest way to close a connection.
1165
1166This method may invoke callbacks (and therefore the handle might be
1167destroyed after it returns).
1168
1169=cut
1170
1171sub push_shutdown {
1172 my ($self) = @_;
1173
1174 delete $self->{low_water_mark};
1175 $self->on_drain (sub { shutdown $_[0]{fh}, 1 });
1176}
1177
1178=item custom write types - Package::anyevent_write_type $handle, @args
1179
1180Instead of one of the predefined types, you can also specify the name of
1181a package. AnyEvent will try to load the package and then expects to find
1182a function named C<anyevent_write_type> inside. If it isn't found, it
1183progressively tries to load the parent package until it either finds the
1184function (good) or runs out of packages (bad).
1185
386Whenever the given C<type> is used, C<push_write> will invoke the code 1186Whenever the given C<type> is used, C<push_write> will the function with
387reference with the handle object and the remaining arguments. 1187the handle object and the remaining arguments.
388 1188
389The code reference is supposed to return a single octet string that will 1189The function is supposed to return a single octet string that will be
390be appended to the write buffer. 1190appended to the write buffer, so you can mentally treat this function as a
1191"arguments to on-the-wire-format" converter.
391 1192
392Note that this is a function, and all types registered this way will be 1193Example: implement a custom write type C<join> that joins the remaining
393global, so try to use unique names. 1194arguments using the first one.
1195
1196 $handle->push_write (My::Type => " ", 1,2,3);
1197
1198 # uses the following package, which can be defined in the "My::Type" or in
1199 # the "My" modules to be auto-loaded, or just about anywhere when the
1200 # My::Type::anyevent_write_type is defined before invoking it.
1201
1202 package My::Type;
1203
1204 sub anyevent_write_type {
1205 my ($handle, $delim, @args) = @_;
1206
1207 join $delim, @args
1208 }
394 1209
395=cut 1210=cut
396 1211
397############################################################################# 1212#############################################################################
398 1213
407ways, the "simple" way, using only C<on_read> and the "complex" way, using 1222ways, the "simple" way, using only C<on_read> and the "complex" way, using
408a queue. 1223a queue.
409 1224
410In the simple case, you just install an C<on_read> callback and whenever 1225In the simple case, you just install an C<on_read> callback and whenever
411new data arrives, it will be called. You can then remove some data (if 1226new data arrives, it will be called. You can then remove some data (if
412enough is there) from the read buffer (C<< $handle->rbuf >>) if you want 1227enough is there) from the read buffer (C<< $handle->rbuf >>). Or you can
413or not. 1228leave the data there if you want to accumulate more (e.g. when only a
1229partial message has been received so far), or change the read queue with
1230e.g. C<push_read>.
414 1231
415In the more complex case, you want to queue multiple callbacks. In this 1232In the more complex case, you want to queue multiple callbacks. In this
416case, AnyEvent::Handle will call the first queued callback each time new 1233case, AnyEvent::Handle will call the first queued callback each time new
417data arrives and removes it when it has done its job (see C<push_read>, 1234data arrives (also the first time it is queued) and remove it when it has
418below). 1235done its job (see C<push_read>, below).
419 1236
420This way you can, for example, push three line-reads, followed by reading 1237This way you can, for example, push three line-reads, followed by reading
421a chunk of data, and AnyEvent::Handle will execute them in order. 1238a chunk of data, and AnyEvent::Handle will execute them in order.
422 1239
423Example 1: EPP protocol parser. EPP sends 4 byte length info, followed by 1240Example 1: EPP protocol parser. EPP sends 4 byte length info, followed by
424the specified number of bytes which give an XML datagram. 1241the specified number of bytes which give an XML datagram.
425 1242
426 # in the default state, expect some header bytes 1243 # in the default state, expect some header bytes
427 $handle->on_read (sub { 1244 $handle->on_read (sub {
428 # some data is here, now queue the length-header-read (4 octets) 1245 # some data is here, now queue the length-header-read (4 octets)
429 shift->unshift_read_chunk (4, sub { 1246 shift->unshift_read (chunk => 4, sub {
430 # header arrived, decode 1247 # header arrived, decode
431 my $len = unpack "N", $_[1]; 1248 my $len = unpack "N", $_[1];
432 1249
433 # now read the payload 1250 # now read the payload
434 shift->unshift_read_chunk ($len, sub { 1251 shift->unshift_read (chunk => $len, sub {
435 my $xml = $_[1]; 1252 my $xml = $_[1];
436 # handle xml 1253 # handle xml
437 }); 1254 });
438 }); 1255 });
439 }); 1256 });
440 1257
441Example 2: Implement a client for a protocol that replies either with 1258Example 2: Implement a client for a protocol that replies either with "OK"
442"OK" and another line or "ERROR" for one request, and 64 bytes for the 1259and another line or "ERROR" for the first request that is sent, and 64
443second request. Due tot he availability of a full queue, we can just 1260bytes for the second request. Due to the availability of a queue, we can
444pipeline sending both requests and manipulate the queue as necessary in 1261just pipeline sending both requests and manipulate the queue as necessary
445the callbacks: 1262in the callbacks.
446 1263
447 # request one 1264When the first callback is called and sees an "OK" response, it will
1265C<unshift> another line-read. This line-read will be queued I<before> the
126664-byte chunk callback.
1267
1268 # request one, returns either "OK + extra line" or "ERROR"
448 $handle->push_write ("request 1\015\012"); 1269 $handle->push_write ("request 1\015\012");
449 1270
450 # we expect "ERROR" or "OK" as response, so push a line read 1271 # we expect "ERROR" or "OK" as response, so push a line read
451 $handle->push_read_line (sub { 1272 $handle->push_read (line => sub {
452 # if we got an "OK", we have to _prepend_ another line, 1273 # if we got an "OK", we have to _prepend_ another line,
453 # so it will be read before the second request reads its 64 bytes 1274 # so it will be read before the second request reads its 64 bytes
454 # which are already in the queue when this callback is called 1275 # which are already in the queue when this callback is called
455 # we don't do this in case we got an error 1276 # we don't do this in case we got an error
456 if ($_[1] eq "OK") { 1277 if ($_[1] eq "OK") {
457 $_[0]->unshift_read_line (sub { 1278 $_[0]->unshift_read (line => sub {
458 my $response = $_[1]; 1279 my $response = $_[1];
459 ... 1280 ...
460 }); 1281 });
461 } 1282 }
462 }); 1283 });
463 1284
464 # request two 1285 # request two, simply returns 64 octets
465 $handle->push_write ("request 2\015\012"); 1286 $handle->push_write ("request 2\015\012");
466 1287
467 # simply read 64 bytes, always 1288 # simply read 64 bytes, always
468 $handle->push_read_chunk (64, sub { 1289 $handle->push_read (chunk => 64, sub {
469 my $response = $_[1]; 1290 my $response = $_[1];
470 ... 1291 ...
471 }); 1292 });
472 1293
473=over 4 1294=over 4
474 1295
475=cut 1296=cut
476 1297
477sub _drain_rbuf { 1298sub _drain_rbuf {
478 my ($self) = @_; 1299 my ($self) = @_;
1300
1301 # avoid recursion
1302 return if $self->{_skip_drain_rbuf};
1303 local $self->{_skip_drain_rbuf} = 1;
1304
1305 while () {
1306 # we need to use a separate tls read buffer, as we must not receive data while
1307 # we are draining the buffer, and this can only happen with TLS.
1308 $self->{rbuf} .= delete $self->{_tls_rbuf}
1309 if exists $self->{_tls_rbuf};
1310
1311 my $len = length $self->{rbuf};
1312
1313 if (my $cb = shift @{ $self->{_queue} }) {
1314 unless ($cb->($self)) {
1315 # no progress can be made
1316 # (not enough data and no data forthcoming)
1317 $self->_error (Errno::EPIPE, 1), return
1318 if $self->{_eof};
1319
1320 unshift @{ $self->{_queue} }, $cb;
1321 last;
1322 }
1323 } elsif ($self->{on_read}) {
1324 last unless $len;
1325
1326 $self->{on_read}($self);
1327
1328 if (
1329 $len == length $self->{rbuf} # if no data has been consumed
1330 && !@{ $self->{_queue} } # and the queue is still empty
1331 && $self->{on_read} # but we still have on_read
1332 ) {
1333 # no further data will arrive
1334 # so no progress can be made
1335 $self->_error (Errno::EPIPE, 1), return
1336 if $self->{_eof};
1337
1338 last; # more data might arrive
1339 }
1340 } else {
1341 # read side becomes idle
1342 delete $self->{_rw} unless $self->{tls};
1343 last;
1344 }
1345 }
1346
1347 if ($self->{_eof}) {
1348 $self->{on_eof}
1349 ? $self->{on_eof}($self)
1350 : $self->_error (0, 1, "Unexpected end-of-file");
1351
1352 return;
1353 }
479 1354
480 if ( 1355 if (
481 defined $self->{rbuf_max} 1356 defined $self->{rbuf_max}
482 && $self->{rbuf_max} < length $self->{rbuf} 1357 && $self->{rbuf_max} < length $self->{rbuf}
483 ) { 1358 ) {
484 $! = &Errno::ENOSPC; 1359 $self->_error (Errno::ENOSPC, 1), return;
485 $self->error;
486 } 1360 }
487 1361
488 return if $self->{in_drain}; 1362 # may need to restart read watcher
489 local $self->{in_drain} = 1; 1363 unless ($self->{_rw}) {
490 1364 $self->start_read
491 while (my $len = length $self->{rbuf}) { 1365 if $self->{on_read} || @{ $self->{_queue} };
492 no strict 'refs';
493 if (my $cb = shift @{ $self->{_queue} }) {
494 unless ($cb->($self)) {
495 if ($self->{_eof}) {
496 # no progress can be made (not enough data and no data forthcoming)
497 $! = &Errno::EPIPE;
498 $self->error;
499 }
500
501 unshift @{ $self->{_queue} }, $cb;
502 return;
503 }
504 } elsif ($self->{on_read}) {
505 $self->{on_read}($self);
506
507 if (
508 $self->{_eof} # if no further data will arrive
509 && $len == length $self->{rbuf} # and no data has been consumed
510 && !@{ $self->{_queue} } # and the queue is still empty
511 && $self->{on_read} # and we still want to read data
512 ) {
513 # then no progress can be made
514 $! = &Errno::EPIPE;
515 $self->error;
516 }
517 } else {
518 # read side becomes idle
519 delete $self->{_rw};
520 return;
521 }
522 }
523
524 if ($self->{_eof}) {
525 $self->_shutdown;
526 $self->{on_eof}($self)
527 if $self->{on_eof};
528 } 1366 }
529} 1367}
530 1368
531=item $handle->on_read ($cb) 1369=item $handle->on_read ($cb)
532 1370
533This replaces the currently set C<on_read> callback, or clears it (when 1371This replaces the currently set C<on_read> callback, or clears it (when
534the new callback is C<undef>). See the description of C<on_read> in the 1372the new callback is C<undef>). See the description of C<on_read> in the
535constructor. 1373constructor.
536 1374
1375This method may invoke callbacks (and therefore the handle might be
1376destroyed after it returns).
1377
537=cut 1378=cut
538 1379
539sub on_read { 1380sub on_read {
540 my ($self, $cb) = @_; 1381 my ($self, $cb) = @_;
541 1382
542 $self->{on_read} = $cb; 1383 $self->{on_read} = $cb;
1384 $self->_drain_rbuf if $cb;
543} 1385}
544 1386
545=item $handle->rbuf 1387=item $handle->rbuf
546 1388
547Returns the read buffer (as a modifiable lvalue). 1389Returns the read buffer (as a modifiable lvalue). You can also access the
1390read buffer directly as the C<< ->{rbuf} >> member, if you want (this is
1391much faster, and no less clean).
548 1392
549You can access the read buffer directly as the C<< ->{rbuf} >> member, if 1393The only operation allowed on the read buffer (apart from looking at it)
550you want. 1394is removing data from its beginning. Otherwise modifying or appending to
1395it is not allowed and will lead to hard-to-track-down bugs.
551 1396
552NOTE: The read buffer should only be used or modified if the C<on_read>, 1397NOTE: The read buffer should only be used or modified in the C<on_read>
553C<push_read> or C<unshift_read> methods are used. The other read methods 1398callback or when C<push_read> or C<unshift_read> are used with a single
554automatically manage the read buffer. 1399callback (i.e. untyped). Typed C<push_read> and C<unshift_read> methods
1400will manage the read buffer on their own.
555 1401
556=cut 1402=cut
557 1403
558sub rbuf : lvalue { 1404sub rbuf : lvalue {
559 $_[0]{rbuf} 1405 $_[0]{rbuf}
576 1422
577If enough data was available, then the callback must remove all data it is 1423If enough data was available, then the callback must remove all data it is
578interested in (which can be none at all) and return a true value. After returning 1424interested in (which can be none at all) and return a true value. After returning
579true, it will be removed from the queue. 1425true, it will be removed from the queue.
580 1426
1427These methods may invoke callbacks (and therefore the handle might be
1428destroyed after it returns).
1429
581=cut 1430=cut
582 1431
583our %RH; 1432our %RH;
584 1433
585sub register_read_type($$) { 1434sub register_read_type($$) {
591 my $cb = pop; 1440 my $cb = pop;
592 1441
593 if (@_) { 1442 if (@_) {
594 my $type = shift; 1443 my $type = shift;
595 1444
1445 $cb = ($RH{$type} ||= _load_func "$type\::anyevent_read_type"
596 $cb = ($RH{$type} or Carp::croak "unsupported type passed to AnyEvent::Handle::push_read") 1446 or Carp::croak "unsupported/unloadable type '$type' passed to AnyEvent::Handle::push_read")
597 ->($self, $cb, @_); 1447 ->($self, $cb, @_);
598 } 1448 }
599 1449
600 push @{ $self->{_queue} }, $cb; 1450 push @{ $self->{_queue} }, $cb;
601 $self->_drain_rbuf; 1451 $self->_drain_rbuf;
606 my $cb = pop; 1456 my $cb = pop;
607 1457
608 if (@_) { 1458 if (@_) {
609 my $type = shift; 1459 my $type = shift;
610 1460
1461 $cb = ($RH{$type} ||= _load_func "$type\::anyevent_read_type"
611 $cb = ($RH{$type} or Carp::croak "unsupported type passed to AnyEvent::Handle::unshift_read") 1462 or Carp::croak "unsupported/unloadable type '$type' passed to AnyEvent::Handle::unshift_read")
612 ->($self, $cb, @_); 1463 ->($self, $cb, @_);
613 } 1464 }
614
615 1465
616 unshift @{ $self->{_queue} }, $cb; 1466 unshift @{ $self->{_queue} }, $cb;
617 $self->_drain_rbuf; 1467 $self->_drain_rbuf;
618} 1468}
619 1469
621 1471
622=item $handle->unshift_read (type => @args, $cb) 1472=item $handle->unshift_read (type => @args, $cb)
623 1473
624Instead of providing a callback that parses the data itself you can chose 1474Instead of providing a callback that parses the data itself you can chose
625between a number of predefined parsing formats, for chunks of data, lines 1475between a number of predefined parsing formats, for chunks of data, lines
626etc. 1476etc. You can also specify the (fully qualified) name of a package, in
1477which case AnyEvent tries to load the package and then expects to find the
1478C<anyevent_read_type> function inside (see "custom read types", below).
627 1479
628Predefined types are (if you have ideas for additional types, feel free to 1480Predefined types are (if you have ideas for additional types, feel free to
629drop by and tell us): 1481drop by and tell us):
630 1482
631=over 4 1483=over 4
632 1484
633=item chunk => $octets, $cb->($self, $data) 1485=item chunk => $octets, $cb->($handle, $data)
634 1486
635Invoke the callback only once C<$octets> bytes have been read. Pass the 1487Invoke the callback only once C<$octets> bytes have been read. Pass the
636data read to the callback. The callback will never be called with less 1488data read to the callback. The callback will never be called with less
637data. 1489data.
638 1490
639Example: read 2 bytes. 1491Example: read 2 bytes.
640 1492
641 $handle->push_read (chunk => 2, sub { 1493 $handle->push_read (chunk => 2, sub {
642 warn "yay ", unpack "H*", $_[1]; 1494 say "yay " . unpack "H*", $_[1];
643 }); 1495 });
644 1496
645=cut 1497=cut
646 1498
647register_read_type chunk => sub { 1499register_read_type chunk => sub {
652 $cb->($_[0], substr $_[0]{rbuf}, 0, $len, ""); 1504 $cb->($_[0], substr $_[0]{rbuf}, 0, $len, "");
653 1 1505 1
654 } 1506 }
655}; 1507};
656 1508
657# compatibility with older API
658sub push_read_chunk {
659 $_[0]->push_read (chunk => $_[1], $_[2]);
660}
661
662sub unshift_read_chunk {
663 $_[0]->unshift_read (chunk => $_[1], $_[2]);
664}
665
666=item line => [$eol, ]$cb->($self, $line, $eol) 1509=item line => [$eol, ]$cb->($handle, $line, $eol)
667 1510
668The callback will be called only once a full line (including the end of 1511The callback will be called only once a full line (including the end of
669line marker, C<$eol>) has been read. This line (excluding the end of line 1512line marker, C<$eol>) has been read. This line (excluding the end of line
670marker) will be passed to the callback as second argument (C<$line>), and 1513marker) will be passed to the callback as second argument (C<$line>), and
671the end of line marker as the third argument (C<$eol>). 1514the end of line marker as the third argument (C<$eol>).
685=cut 1528=cut
686 1529
687register_read_type line => sub { 1530register_read_type line => sub {
688 my ($self, $cb, $eol) = @_; 1531 my ($self, $cb, $eol) = @_;
689 1532
690 $eol = qr|(\015?\012)| if @_ < 3; 1533 if (@_ < 3) {
1534 # this is faster then the generic code below
1535 sub {
1536 (my $pos = index $_[0]{rbuf}, "\012") >= 0
1537 or return;
1538
1539 (my $str = substr $_[0]{rbuf}, 0, $pos + 1, "") =~ s/(\015?\012)\Z// or die;
1540 $cb->($_[0], $str, "$1");
1541 1
1542 }
1543 } else {
691 $eol = quotemeta $eol unless ref $eol; 1544 $eol = quotemeta $eol unless ref $eol;
692 $eol = qr|^(.*?)($eol)|s; 1545 $eol = qr|^(.*?)($eol)|s;
693 1546
694 sub { 1547 sub {
695 $_[0]{rbuf} =~ s/$eol// or return; 1548 $_[0]{rbuf} =~ s/$eol// or return;
696 1549
697 $cb->($_[0], $1, $2); 1550 $cb->($_[0], "$1", "$2");
1551 1
698 1 1552 }
699 } 1553 }
700}; 1554};
701 1555
702# compatibility with older API
703sub push_read_line {
704 my $self = shift;
705 $self->push_read (line => @_);
706}
707
708sub unshift_read_line {
709 my $self = shift;
710 $self->unshift_read (line => @_);
711}
712
713=item netstring => $cb->($string)
714
715A netstring (http://cr.yp.to/proto/netstrings.txt, this is not an endorsement).
716
717Throws an error with C<$!> set to EBADMSG on format violations.
718
719=cut
720
721register_read_type netstring => sub {
722 my ($self, $cb) = @_;
723
724 sub {
725 unless ($_[0]{rbuf} =~ s/^(0|[1-9][0-9]*)://) {
726 if ($_[0]{rbuf} =~ /[^0-9]/) {
727 $! = &Errno::EBADMSG;
728 $self->error;
729 }
730 return;
731 }
732
733 my $len = $1;
734
735 $self->unshift_read (chunk => $len, sub {
736 my $string = $_[1];
737 $_[0]->unshift_read (chunk => 1, sub {
738 if ($_[1] eq ",") {
739 $cb->($_[0], $string);
740 } else {
741 $! = &Errno::EBADMSG;
742 $self->error;
743 }
744 });
745 });
746
747 1
748 }
749};
750
751=item regex => $accept[, $reject[, $skip], $cb->($data) 1556=item regex => $accept[, $reject[, $skip], $cb->($handle, $data)
752 1557
753Makes a regex match against the regex object C<$accept> and returns 1558Makes a regex match against the regex object C<$accept> and returns
754everything up to and including the match. 1559everything up to and including the match.
755 1560
756Example: read a single line terminated by '\n'. 1561Example: read a single line terminated by '\n'.
772the receive buffer when neither C<$accept> nor C<$reject> match, 1577the receive buffer when neither C<$accept> nor C<$reject> match,
773and everything preceding and including the match will be accepted 1578and everything preceding and including the match will be accepted
774unconditionally. This is useful to skip large amounts of data that you 1579unconditionally. This is useful to skip large amounts of data that you
775know cannot be matched, so that the C<$accept> or C<$reject> regex do not 1580know cannot be matched, so that the C<$accept> or C<$reject> regex do not
776have to start matching from the beginning. This is purely an optimisation 1581have to start matching from the beginning. This is purely an optimisation
777and is usually worth only when you expect more than a few kilobytes. 1582and is usually worth it only when you expect more than a few kilobytes.
778 1583
779Example: expect a http header, which ends at C<\015\012\015\012>. Since we 1584Example: expect a http header, which ends at C<\015\012\015\012>. Since we
780expect the header to be very large (it isn't in practise, but...), we use 1585expect the header to be very large (it isn't in practice, but...), we use
781a skip regex to skip initial portions. The skip regex is tricky in that 1586a skip regex to skip initial portions. The skip regex is tricky in that
782it only accepts something not ending in either \015 or \012, as these are 1587it only accepts something not ending in either \015 or \012, as these are
783required for the accept regex. 1588required for the accept regex.
784 1589
785 $handle->push_read (regex => 1590 $handle->push_read (regex =>
798 1603
799 sub { 1604 sub {
800 # accept 1605 # accept
801 if ($$rbuf =~ $accept) { 1606 if ($$rbuf =~ $accept) {
802 $data .= substr $$rbuf, 0, $+[0], ""; 1607 $data .= substr $$rbuf, 0, $+[0], "";
803 $cb->($self, $data); 1608 $cb->($_[0], $data);
804 return 1; 1609 return 1;
805 } 1610 }
806 1611
807 # reject 1612 # reject
808 if ($reject && $$rbuf =~ $reject) { 1613 if ($reject && $$rbuf =~ $reject) {
809 $! = &Errno::EBADMSG; 1614 $_[0]->_error (Errno::EBADMSG);
810 $self->error;
811 } 1615 }
812 1616
813 # skip 1617 # skip
814 if ($skip && $$rbuf =~ $skip) { 1618 if ($skip && $$rbuf =~ $skip) {
815 $data .= substr $$rbuf, 0, $+[0], ""; 1619 $data .= substr $$rbuf, 0, $+[0], "";
817 1621
818 () 1622 ()
819 } 1623 }
820}; 1624};
821 1625
1626=item netstring => $cb->($handle, $string)
1627
1628A netstring (http://cr.yp.to/proto/netstrings.txt, this is not an endorsement).
1629
1630Throws an error with C<$!> set to EBADMSG on format violations.
1631
1632=cut
1633
1634register_read_type netstring => sub {
1635 my ($self, $cb) = @_;
1636
1637 sub {
1638 unless ($_[0]{rbuf} =~ s/^(0|[1-9][0-9]*)://) {
1639 if ($_[0]{rbuf} =~ /[^0-9]/) {
1640 $_[0]->_error (Errno::EBADMSG);
1641 }
1642 return;
1643 }
1644
1645 my $len = $1;
1646
1647 $_[0]->unshift_read (chunk => $len, sub {
1648 my $string = $_[1];
1649 $_[0]->unshift_read (chunk => 1, sub {
1650 if ($_[1] eq ",") {
1651 $cb->($_[0], $string);
1652 } else {
1653 $_[0]->_error (Errno::EBADMSG);
1654 }
1655 });
1656 });
1657
1658 1
1659 }
1660};
1661
1662=item packstring => $format, $cb->($handle, $string)
1663
1664An octet string prefixed with an encoded length. The encoding C<$format>
1665uses the same format as a Perl C<pack> format, but must specify a single
1666integer only (only one of C<cCsSlLqQiInNvVjJw> is allowed, plus an
1667optional C<!>, C<< < >> or C<< > >> modifier).
1668
1669For example, DNS over TCP uses a prefix of C<n> (2 octet network order),
1670EPP uses a prefix of C<N> (4 octtes).
1671
1672Example: read a block of data prefixed by its length in BER-encoded
1673format (very efficient).
1674
1675 $handle->push_read (packstring => "w", sub {
1676 my ($handle, $data) = @_;
1677 });
1678
1679=cut
1680
1681register_read_type packstring => sub {
1682 my ($self, $cb, $format) = @_;
1683
1684 sub {
1685 # when we can use 5.10 we can use ".", but for 5.8 we use the re-pack method
1686 defined (my $len = eval { unpack $format, $_[0]{rbuf} })
1687 or return;
1688
1689 $format = length pack $format, $len;
1690
1691 # bypass unshift if we already have the remaining chunk
1692 if ($format + $len <= length $_[0]{rbuf}) {
1693 my $data = substr $_[0]{rbuf}, $format, $len;
1694 substr $_[0]{rbuf}, 0, $format + $len, "";
1695 $cb->($_[0], $data);
1696 } else {
1697 # remove prefix
1698 substr $_[0]{rbuf}, 0, $format, "";
1699
1700 # read remaining chunk
1701 $_[0]->unshift_read (chunk => $len, $cb);
1702 }
1703
1704 1
1705 }
1706};
1707
1708=item json => $cb->($handle, $hash_or_arrayref)
1709
1710Reads a JSON object or array, decodes it and passes it to the
1711callback. When a parse error occurs, an C<EBADMSG> error will be raised.
1712
1713If a C<json> object was passed to the constructor, then that will be used
1714for the final decode, otherwise it will create a JSON coder expecting UTF-8.
1715
1716This read type uses the incremental parser available with JSON version
17172.09 (and JSON::XS version 2.2) and above. You have to provide a
1718dependency on your own: this module will load the JSON module, but
1719AnyEvent does not depend on it itself.
1720
1721Since JSON texts are fully self-delimiting, the C<json> read and write
1722types are an ideal simple RPC protocol: just exchange JSON datagrams. See
1723the C<json> write type description, above, for an actual example.
1724
1725=cut
1726
1727register_read_type json => sub {
1728 my ($self, $cb) = @_;
1729
1730 my $json = $self->{json} ||= json_coder;
1731
1732 my $data;
1733
1734 sub {
1735 my $ref = eval { $json->incr_parse ($_[0]{rbuf}) };
1736
1737 if ($ref) {
1738 $_[0]{rbuf} = $json->incr_text;
1739 $json->incr_text = "";
1740 $cb->($_[0], $ref);
1741
1742 1
1743 } elsif ($@) {
1744 # error case
1745 $json->incr_skip;
1746
1747 $_[0]{rbuf} = $json->incr_text;
1748 $json->incr_text = "";
1749
1750 $_[0]->_error (Errno::EBADMSG);
1751
1752 ()
1753 } else {
1754 $_[0]{rbuf} = "";
1755
1756 ()
1757 }
1758 }
1759};
1760
1761=item cbor => $cb->($handle, $scalar)
1762
1763Reads a CBOR value, decodes it and passes it to the callback. When a parse
1764error occurs, an C<EBADMSG> error will be raised.
1765
1766If a L<CBOR::XS> object was passed to the constructor, then that will be
1767used for the final decode, otherwise it will create a CBOR coder without
1768enabling any options.
1769
1770You have to provide a dependency to L<CBOR::XS> on your own: this module
1771will load the L<CBOR::XS> module, but AnyEvent does not depend on it
1772itself.
1773
1774Since CBOR values are fully self-delimiting, the C<cbor> read and write
1775types are an ideal simple RPC protocol: just exchange CBOR datagrams. See
1776the C<cbor> write type description, above, for an actual example.
1777
1778=cut
1779
1780register_read_type cbor => sub {
1781 my ($self, $cb) = @_;
1782
1783 my $cbor = $self->{cbor} ||= cbor_coder;
1784
1785 my $data;
1786
1787 sub {
1788 my (@value) = eval { $cbor->incr_parse ($_[0]{rbuf}) };
1789
1790 if (@value) {
1791 $cb->($_[0], @value);
1792
1793 1
1794 } elsif ($@) {
1795 # error case
1796 $cbor->incr_reset;
1797
1798 $_[0]->_error (Errno::EBADMSG);
1799
1800 ()
1801 } else {
1802 ()
1803 }
1804 }
1805};
1806
1807=item storable => $cb->($handle, $ref)
1808
1809Deserialises a L<Storable> frozen representation as written by the
1810C<storable> write type (BER-encoded length prefix followed by nfreeze'd
1811data).
1812
1813Raises C<EBADMSG> error if the data could not be decoded.
1814
1815=cut
1816
1817register_read_type storable => sub {
1818 my ($self, $cb) = @_;
1819
1820 require Storable unless $Storable::VERSION;
1821
1822 sub {
1823 # when we can use 5.10 we can use ".", but for 5.8 we use the re-pack method
1824 defined (my $len = eval { unpack "w", $_[0]{rbuf} })
1825 or return;
1826
1827 my $format = length pack "w", $len;
1828
1829 # bypass unshift if we already have the remaining chunk
1830 if ($format + $len <= length $_[0]{rbuf}) {
1831 my $data = substr $_[0]{rbuf}, $format, $len;
1832 substr $_[0]{rbuf}, 0, $format + $len, "";
1833
1834 eval { $cb->($_[0], Storable::thaw ($data)); 1 }
1835 or return $_[0]->_error (Errno::EBADMSG);
1836 } else {
1837 # remove prefix
1838 substr $_[0]{rbuf}, 0, $format, "";
1839
1840 # read remaining chunk
1841 $_[0]->unshift_read (chunk => $len, sub {
1842 eval { $cb->($_[0], Storable::thaw ($_[1])); 1 }
1843 or $_[0]->_error (Errno::EBADMSG);
1844 });
1845 }
1846
1847 1
1848 }
1849};
1850
1851=item tls_detect => $cb->($handle, $detect, $major, $minor)
1852
1853Checks the input stream for a valid SSL or TLS handshake TLSPaintext
1854record without consuming anything. Only SSL version 3 or higher
1855is handled, up to the fictituous protocol 4.x (but both SSL3+ and
1856SSL2-compatible framing is supported).
1857
1858If it detects that the input data is likely TLS, it calls the callback
1859with a true value for C<$detect> and the (on-wire) TLS version as second
1860and third argument (C<$major> is C<3>, and C<$minor> is 0..3 for SSL
18613.0, TLS 1.0, 1.1 and 1.2, respectively). If it detects the input to
1862be definitely not TLS, it calls the callback with a false value for
1863C<$detect>.
1864
1865The callback could use this information to decide whether or not to start
1866TLS negotiation.
1867
1868In all cases the data read so far is passed to the following read
1869handlers.
1870
1871Usually you want to use the C<tls_autostart> read type instead.
1872
1873If you want to design a protocol that works in the presence of TLS
1874dtection, make sure that any non-TLS data doesn't start with the octet 22
1875(ASCII SYN, 16 hex) or 128-255 (i.e. highest bit set). The checks this
1876read type does are a bit more strict, but might losen in the future to
1877accomodate protocol changes.
1878
1879This read type does not rely on L<AnyEvent::TLS> (and thus, not on
1880L<Net::SSLeay>).
1881
1882=item tls_autostart => $tls[, $tls_ctx]
1883
1884Tries to detect a valid SSL or TLS handshake. If one is detected, it tries
1885to start tls by calling C<starttls> with the given arguments.
1886
1887In practise, C<$tls> must be C<accept>, or a Net::SSLeay context that has
1888been configured to accept, as servers do not normally send a handshake on
1889their own and ths cannot be detected in this way.
1890
1891See C<tls_detect> above for more details.
1892
1893Example: give the client a chance to start TLS before accepting a text
1894line.
1895
1896 $hdl->push_read (tls_detect => "accept");
1897 $hdl->push_read (line => sub {
1898 print "received ", ($_[0]{tls} ? "encrypted" : "cleartext"), " <$_[1]>\n";
1899 });
1900
1901=cut
1902
1903register_read_type tls_detect => sub {
1904 my ($self, $cb) = @_;
1905
1906 sub {
1907 # this regex matches a full or partial tls record
1908 if (
1909 # ssl3+: type(22=handshake) major(=3) minor(any) length_hi
1910 $self->{rbuf} =~ /^(?:\z| \x16 (\z| [\x03\x04] (?:\z| . (?:\z| [\x00-\x40] ))))/xs
1911 # ssl2 comapatible: len_hi len_lo type(1) major minor dummy(forlength)
1912 or $self->{rbuf} =~ /^(?:\z| [\x80-\xff] (?:\z| . (?:\z| \x01 (\z| [\x03\x04] (?:\z| . (?:\z| . ))))))/xs
1913 ) {
1914 return if 3 != length $1; # partial match, can't decide yet
1915
1916 # full match, valid TLS record
1917 my ($major, $minor) = unpack "CC", $1;
1918 $cb->($self, "accept", $major + $minor * 0.1);
1919 } else {
1920 # mismatch == guaranteed not TLS
1921 $cb->($self, undef);
1922 }
1923
1924 1
1925 }
1926};
1927
1928register_read_type tls_autostart => sub {
1929 my ($self, @tls) = @_;
1930
1931 $RH{tls_detect}($self, sub {
1932 return unless $_[1];
1933 $_[0]->starttls (@tls);
1934 })
1935};
1936
822=back 1937=back
823 1938
824=item AnyEvent::Handle::register_read_type type => $coderef->($self, $cb, @args) 1939=item custom read types - Package::anyevent_read_type $handle, $cb, @args
825 1940
826This function (not method) lets you add your own types to C<push_read>. 1941Instead of one of the predefined types, you can also specify the name
1942of a package. AnyEvent will try to load the package and then expects to
1943find a function named C<anyevent_read_type> inside. If it isn't found, it
1944progressively tries to load the parent package until it either finds the
1945function (good) or runs out of packages (bad).
827 1946
828Whenever the given C<type> is used, C<push_read> will invoke the code 1947Whenever this type is used, C<push_read> will invoke the function with the
829reference with the handle object, the callback and the remaining 1948handle object, the original callback and the remaining arguments.
830arguments.
831 1949
832The code reference is supposed to return a callback (usually a closure) 1950The function is supposed to return a callback (usually a closure) that
833that works as a plain read callback (see C<< ->push_read ($cb) >>). 1951works as a plain read callback (see C<< ->push_read ($cb) >>), so you can
1952mentally treat the function as a "configurable read type to read callback"
1953converter.
834 1954
835It should invoke the passed callback when it is done reading (remember to 1955It should invoke the original callback when it is done reading (remember
836pass C<$self> as first argument as all other callbacks do that). 1956to pass C<$handle> as first argument as all other callbacks do that,
1957although there is no strict requirement on this).
837 1958
838Note that this is a function, and all types registered this way will be
839global, so try to use unique names.
840
841For examples, see the source of this module (F<perldoc -m AnyEvent::Handle>, 1959For examples, see the source of this module (F<perldoc -m
842search for C<register_read_type>)). 1960AnyEvent::Handle>, search for C<register_read_type>)).
843 1961
844=item $handle->stop_read 1962=item $handle->stop_read
845 1963
846=item $handle->start_read 1964=item $handle->start_read
847 1965
848In rare cases you actually do not want to read anything from the 1966In rare cases you actually do not want to read anything from the
849socket. In this case you can call C<stop_read>. Neither C<on_read> no 1967socket. In this case you can call C<stop_read>. Neither C<on_read> nor
850any queued callbacks will be executed then. To start reading again, call 1968any queued callbacks will be executed then. To start reading again, call
851C<start_read>. 1969C<start_read>.
852 1970
1971Note that AnyEvent::Handle will automatically C<start_read> for you when
1972you change the C<on_read> callback or push/unshift a read callback, and it
1973will automatically C<stop_read> for you when neither C<on_read> is set nor
1974there are any read requests in the queue.
1975
1976In older versions of this module (<= 5.3), these methods had no effect,
1977as TLS does not support half-duplex connections. In current versions they
1978work as expected, as this behaviour is required to avoid certain resource
1979attacks, where the program would be forced to read (and buffer) arbitrary
1980amounts of data before being able to send some data. The drawback is that
1981some readings of the the SSL/TLS specifications basically require this
1982attack to be working, as SSL/TLS implementations might stall sending data
1983during a rehandshake.
1984
1985As a guideline, during the initial handshake, you should not stop reading,
1986and as a client, it might cause problems, depending on your application.
1987
853=cut 1988=cut
854 1989
855sub stop_read { 1990sub stop_read {
856 my ($self) = @_; 1991 my ($self) = @_;
857 1992
859} 1994}
860 1995
861sub start_read { 1996sub start_read {
862 my ($self) = @_; 1997 my ($self) = @_;
863 1998
864 unless ($self->{_rw} || $self->{_eof}) { 1999 unless ($self->{_rw} || $self->{_eof} || !$self->{fh}) {
865 Scalar::Util::weaken $self; 2000 Scalar::Util::weaken $self;
866 2001
867 $self->{_rw} = AnyEvent->io (fh => $self->{fh}, poll => "r", cb => sub { 2002 $self->{_rw} = AE::io $self->{fh}, 0, sub {
868 my $rbuf = $self->{filter_r} ? \my $buf : \$self->{rbuf}; 2003 my $rbuf = \($self->{tls} ? my $buf : $self->{rbuf});
869 my $len = sysread $self->{fh}, $$rbuf, $self->{read_size} || 8192, length $$rbuf; 2004 my $len = sysread $self->{fh}, $$rbuf, $self->{read_size}, length $$rbuf;
870 2005
871 if ($len > 0) { 2006 if ($len > 0) {
872 $self->{filter_r} 2007 $self->{_activity} = $self->{_ractivity} = AE::now;
873 ? $self->{filter_r}->($self, $rbuf) 2008
2009 if ($self->{tls}) {
2010 Net::SSLeay::BIO_write ($self->{_rbio}, $$rbuf);
2011
2012 &_dotls ($self);
2013 } else {
874 : $self->_drain_rbuf; 2014 $self->_drain_rbuf;
2015 }
2016
2017 if ($len == $self->{read_size}) {
2018 $self->{read_size} *= 2;
2019 $self->{read_size} = $self->{max_read_size} || MAX_READ_SIZE
2020 if $self->{read_size} > ($self->{max_read_size} || MAX_READ_SIZE);
2021 }
875 2022
876 } elsif (defined $len) { 2023 } elsif (defined $len) {
877 delete $self->{_rw}; 2024 delete $self->{_rw};
878 $self->{_eof} = 1; 2025 $self->{_eof} = 1;
879 $self->_drain_rbuf; 2026 $self->_drain_rbuf;
880 2027
881 } elsif ($! != EAGAIN && $! != EINTR && $! != &AnyEvent::Util::WSAWOULDBLOCK) { 2028 } elsif ($! != EAGAIN && $! != EINTR && $! != WSAEWOULDBLOCK) {
882 return $self->error; 2029 return $self->_error ($!, 1);
883 } 2030 }
884 }); 2031 };
885 } 2032 }
886} 2033}
887 2034
2035our $ERROR_SYSCALL;
2036our $ERROR_WANT_READ;
2037
2038sub _tls_error {
2039 my ($self, $err) = @_;
2040
2041 return $self->_error ($!, 1)
2042 if $err == Net::SSLeay::ERROR_SYSCALL ();
2043
2044 my $err = Net::SSLeay::ERR_error_string (Net::SSLeay::ERR_get_error ());
2045
2046 # reduce error string to look less scary
2047 $err =~ s/^error:[0-9a-fA-F]{8}:[^:]+:([^:]+):/\L$1: /;
2048
2049 if ($self->{_on_starttls}) {
2050 (delete $self->{_on_starttls})->($self, undef, $err);
2051 &_freetls;
2052 } else {
2053 &_freetls;
2054 $self->_error (Errno::EPROTO, 1, $err);
2055 }
2056}
2057
2058# poll the write BIO and send the data if applicable
2059# also decode read data if possible
2060# this is basiclaly our TLS state machine
2061# more efficient implementations are possible with openssl,
2062# but not with the buggy and incomplete Net::SSLeay.
888sub _dotls { 2063sub _dotls {
889 my ($self) = @_; 2064 my ($self) = @_;
890 2065
2066 my $tmp;
2067
891 if (length $self->{_tls_wbuf}) { 2068 while (length $self->{_tls_wbuf}) {
892 while ((my $len = Net::SSLeay::write ($self->{tls}, $self->{_tls_wbuf})) > 0) { 2069 if (($tmp = Net::SSLeay::write ($self->{tls}, $self->{_tls_wbuf})) <= 0) {
893 substr $self->{_tls_wbuf}, 0, $len, ""; 2070 $tmp = Net::SSLeay::get_error ($self->{tls}, $tmp);
2071
2072 return $self->_tls_error ($tmp)
2073 if $tmp != $ERROR_WANT_READ
2074 && ($tmp != $ERROR_SYSCALL || $!);
2075
2076 last;
894 } 2077 }
895 }
896 2078
2079 substr $self->{_tls_wbuf}, 0, $tmp, "";
2080 }
2081
2082 while (defined ($tmp = Net::SSLeay::read ($self->{tls}))) {
2083 unless (length $tmp) {
2084 $self->{_on_starttls}
2085 and (delete $self->{_on_starttls})->($self, undef, "EOF during handshake"); # ???
2086 &_freetls;
2087
2088 if ($self->{on_stoptls}) {
2089 $self->{on_stoptls}($self);
2090 return;
2091 } else {
2092 # let's treat SSL-eof as we treat normal EOF
2093 delete $self->{_rw};
2094 $self->{_eof} = 1;
2095 }
2096 }
2097
2098 $self->{_tls_rbuf} .= $tmp;
2099 $self->_drain_rbuf;
2100 $self->{tls} or return; # tls session might have gone away in callback
2101 }
2102
2103 $tmp = Net::SSLeay::get_error ($self->{tls}, -1); # -1 is not neccessarily correct, but Net::SSLeay doesn't tell us
2104 return $self->_tls_error ($tmp)
2105 if $tmp != $ERROR_WANT_READ
2106 && ($tmp != $ERROR_SYSCALL || $!);
2107
897 if (defined (my $buf = Net::SSLeay::BIO_read ($self->{_wbio}))) { 2108 while (length ($tmp = Net::SSLeay::BIO_read ($self->{_wbio}))) {
898 $self->{wbuf} .= $buf; 2109 $self->{wbuf} .= $tmp;
899 $self->_drain_wbuf; 2110 $self->_drain_wbuf;
2111 $self->{tls} or return; # tls session might have gone away in callback
900 } 2112 }
901 2113
902 while (defined (my $buf = Net::SSLeay::read ($self->{tls}))) { 2114 $self->{_on_starttls}
903 $self->{rbuf} .= $buf; 2115 and Net::SSLeay::state ($self->{tls}) == Net::SSLeay::ST_OK ()
904 $self->_drain_rbuf; 2116 and (delete $self->{_on_starttls})->($self, 1, "TLS/SSL connection established");
905 }
906
907 my $err = Net::SSLeay::get_error ($self->{tls}, -1);
908
909 if ($err!= Net::SSLeay::ERROR_WANT_READ ()) {
910 if ($err == Net::SSLeay::ERROR_SYSCALL ()) {
911 $self->error;
912 } elsif ($err == Net::SSLeay::ERROR_SSL ()) {
913 $! = &Errno::EIO;
914 $self->error;
915 }
916
917 # all others are fine for our purposes
918 }
919} 2117}
920 2118
921=item $handle->starttls ($tls[, $tls_ctx]) 2119=item $handle->starttls ($tls[, $tls_ctx])
922 2120
923Instead of starting TLS negotiation immediately when the AnyEvent::Handle 2121Instead of starting TLS negotiation immediately when the AnyEvent::Handle
924object is created, you can also do that at a later time by calling 2122object is created, you can also do that at a later time by calling
925C<starttls>. 2123C<starttls>. See the C<tls> constructor argument for general info.
2124
2125Starting TLS is currently an asynchronous operation - when you push some
2126write data and then call C<< ->starttls >> then TLS negotiation will start
2127immediately, after which the queued write data is then sent. This might
2128change in future versions, so best make sure you have no outstanding write
2129data when calling this method.
926 2130
927The first argument is the same as the C<tls> constructor argument (either 2131The first argument is the same as the C<tls> constructor argument (either
928C<"connect">, C<"accept"> or an existing Net::SSLeay object). 2132C<"connect">, C<"accept"> or an existing Net::SSLeay object).
929 2133
930The second argument is the optional C<Net::SSLeay::CTX> object that is 2134The second argument is the optional C<AnyEvent::TLS> object that is used
931used when AnyEvent::Handle has to create its own TLS connection object. 2135when AnyEvent::Handle has to create its own TLS connection object, or
2136a hash reference with C<< key => value >> pairs that will be used to
2137construct a new context.
932 2138
933The TLS connection object will end up in C<< $handle->{tls} >> after this 2139The TLS connection object will end up in C<< $handle->{tls} >>, the TLS
934call and can be used or changed to your liking. Note that the handshake 2140context in C<< $handle->{tls_ctx} >> after this call and can be used or
935might have already started when this function returns. 2141changed to your liking. Note that the handshake might have already started
2142when this function returns.
936 2143
937=cut 2144Due to bugs in OpenSSL, it might or might not be possible to do multiple
2145handshakes on the same stream. It is best to not attempt to use the
2146stream after stopping TLS.
938 2147
939# TODO: maybe document... 2148This method may invoke callbacks (and therefore the handle might be
2149destroyed after it returns).
2150
2151=cut
2152
2153our %TLS_CACHE; #TODO not yet documented, should we?
2154
940sub starttls { 2155sub starttls {
941 my ($self, $ssl, $ctx) = @_; 2156 my ($self, $tls, $ctx) = @_;
942 2157
943 $self->stoptls; 2158 Carp::croak "It is an error to call starttls on an AnyEvent::Handle object while TLS is already active, caught"
2159 if $self->{tls};
944 2160
945 if ($ssl eq "accept") { 2161 unless (defined $AnyEvent::TLS::VERSION) {
946 $ssl = Net::SSLeay::new ($ctx || TLS_CTX ()); 2162 eval {
947 Net::SSLeay::set_accept_state ($ssl); 2163 require Net::SSLeay;
948 } elsif ($ssl eq "connect") { 2164 require AnyEvent::TLS;
949 $ssl = Net::SSLeay::new ($ctx || TLS_CTX ()); 2165 1
950 Net::SSLeay::set_connect_state ($ssl); 2166 } or return $self->_error (Errno::EPROTO, 1, "TLS support not available on this system");
2167 }
2168
2169 $self->{tls} = $tls;
2170 $self->{tls_ctx} = $ctx if @_ > 2;
2171
2172 return unless $self->{fh};
2173
2174 $ERROR_SYSCALL = Net::SSLeay::ERROR_SYSCALL ();
2175 $ERROR_WANT_READ = Net::SSLeay::ERROR_WANT_READ ();
2176
2177 $tls = delete $self->{tls};
2178 $ctx = $self->{tls_ctx};
2179
2180 local $Carp::CarpLevel = 1; # skip ourselves when creating a new context or session
2181
2182 if ("HASH" eq ref $ctx) {
2183 if ($ctx->{cache}) {
2184 my $key = $ctx+0;
2185 $ctx = $TLS_CACHE{$key} ||= new AnyEvent::TLS %$ctx;
2186 } else {
2187 $ctx = new AnyEvent::TLS %$ctx;
2188 }
2189 }
951 } 2190
952 2191 $self->{tls_ctx} = $ctx || TLS_CTX ();
953 $self->{tls} = $ssl; 2192 $self->{tls} = $tls = $self->{tls_ctx}->_get_session ($tls, $self, $self->{peername});
954 2193
955 # basically, this is deep magic (because SSL_read should have the same issues) 2194 # basically, this is deep magic (because SSL_read should have the same issues)
956 # but the openssl maintainers basically said: "trust us, it just works". 2195 # but the openssl maintainers basically said: "trust us, it just works".
957 # (unfortunately, we have to hardcode constants because the abysmally misdesigned 2196 # (unfortunately, we have to hardcode constants because the abysmally misdesigned
958 # and mismaintained ssleay-module doesn't even offer them). 2197 # and mismaintained ssleay-module doesn't even offer them).
959 # http://www.mail-archive.com/openssl-dev@openssl.org/msg22420.html 2198 # http://www.mail-archive.com/openssl-dev@openssl.org/msg22420.html
2199 #
2200 # in short: this is a mess.
2201 #
2202 # note that we do not try to keep the length constant between writes as we are required to do.
2203 # we assume that most (but not all) of this insanity only applies to non-blocking cases,
2204 # and we drive openssl fully in blocking mode here. Or maybe we don't - openssl seems to
2205 # have identity issues in that area.
960 Net::SSLeay::CTX_set_mode ($self->{tls}, 2206# Net::SSLeay::CTX_set_mode ($ssl,
961 (eval { local $SIG{__DIE__}; Net::SSLeay::MODE_ENABLE_PARTIAL_WRITE () } || 1) 2207# (eval { local $SIG{__DIE__}; Net::SSLeay::MODE_ENABLE_PARTIAL_WRITE () } || 1)
962 | (eval { local $SIG{__DIE__}; Net::SSLeay::MODE_ACCEPT_MOVING_WRITE_BUFFER () } || 2)); 2208# | (eval { local $SIG{__DIE__}; Net::SSLeay::MODE_ACCEPT_MOVING_WRITE_BUFFER () } || 2));
2209 Net::SSLeay::CTX_set_mode ($tls, 1|2);
963 2210
964 $self->{_rbio} = Net::SSLeay::BIO_new (Net::SSLeay::BIO_s_mem ()); 2211 $self->{_rbio} = Net::SSLeay::BIO_new (Net::SSLeay::BIO_s_mem ());
965 $self->{_wbio} = Net::SSLeay::BIO_new (Net::SSLeay::BIO_s_mem ()); 2212 $self->{_wbio} = Net::SSLeay::BIO_new (Net::SSLeay::BIO_s_mem ());
966 2213
2214 Net::SSLeay::BIO_write ($self->{_rbio}, $self->{rbuf});
2215 $self->{rbuf} = "";
2216
967 Net::SSLeay::set_bio ($ssl, $self->{_rbio}, $self->{_wbio}); 2217 Net::SSLeay::set_bio ($tls, $self->{_rbio}, $self->{_wbio});
968 2218
969 $self->{filter_w} = sub { 2219 $self->{_on_starttls} = sub { $_[0]{on_starttls}(@_) }
970 $_[0]{_tls_wbuf} .= ${$_[1]}; 2220 if $self->{on_starttls};
971 &_dotls; 2221
972 }; 2222 &_dotls; # need to trigger the initial handshake
973 $self->{filter_r} = sub { 2223 $self->start_read; # make sure we actually do read
974 Net::SSLeay::BIO_write ($_[0]{_rbio}, ${$_[1]});
975 &_dotls;
976 };
977} 2224}
978 2225
979=item $handle->stoptls 2226=item $handle->stoptls
980 2227
981Destroys the SSL connection, if any. Partial read or write data will be 2228Shuts down the SSL connection - this makes a proper EOF handshake by
982lost. 2229sending a close notify to the other side, but since OpenSSL doesn't
2230support non-blocking shut downs, it is not guaranteed that you can re-use
2231the stream afterwards.
2232
2233This method may invoke callbacks (and therefore the handle might be
2234destroyed after it returns).
983 2235
984=cut 2236=cut
985 2237
986sub stoptls { 2238sub stoptls {
987 my ($self) = @_; 2239 my ($self) = @_;
988 2240
989 Net::SSLeay::free (delete $self->{tls}) if $self->{tls}; 2241 if ($self->{tls} && $self->{fh}) {
2242 Net::SSLeay::shutdown ($self->{tls});
990 2243
991 delete $self->{_rbio}; 2244 &_dotls;
992 delete $self->{_wbio}; 2245
993 delete $self->{_tls_wbuf}; 2246# # we don't give a shit. no, we do, but we can't. no...#d#
994 delete $self->{filter_r}; 2247# # we, we... have to use openssl :/#d#
995 delete $self->{filter_w}; 2248# &_freetls;#d#
2249 }
996} 2250}
2251
2252sub _freetls {
2253 my ($self) = @_;
2254
2255 return unless $self->{tls};
2256
2257 $self->{tls_ctx}->_put_session (delete $self->{tls})
2258 if $self->{tls} > 0;
2259
2260 delete @$self{qw(_rbio _wbio _tls_wbuf _on_starttls)};
2261}
2262
2263=item $handle->resettls
2264
2265This rarely-used method simply resets and TLS state on the handle, usually
2266causing data loss.
2267
2268One case where it may be useful is when you want to skip over the data in
2269the stream but you are not interested in interpreting it, so data loss is
2270no concern.
2271
2272=cut
2273
2274*resettls = \&_freetls;
997 2275
998sub DESTROY { 2276sub DESTROY {
999 my $self = shift; 2277 my ($self) = @_;
1000 2278
1001 $self->stoptls; 2279 &_freetls;
2280
2281 my $linger = exists $self->{linger} ? $self->{linger} : 3600;
2282
2283 if ($linger && length $self->{wbuf} && $self->{fh}) {
2284 my $fh = delete $self->{fh};
2285 my $wbuf = delete $self->{wbuf};
2286
2287 my @linger;
2288
2289 push @linger, AE::io $fh, 1, sub {
2290 my $len = syswrite $fh, $wbuf, length $wbuf;
2291
2292 if ($len > 0) {
2293 substr $wbuf, 0, $len, "";
2294 } elsif (defined $len || ($! != EAGAIN && $! != EINTR && $! != WSAEWOULDBLOCK)) {
2295 @linger = (); # end
2296 }
2297 };
2298 push @linger, AE::timer $linger, 0, sub {
2299 @linger = ();
2300 };
2301 }
1002} 2302}
2303
2304=item $handle->destroy
2305
2306Shuts down the handle object as much as possible - this call ensures that
2307no further callbacks will be invoked and as many resources as possible
2308will be freed. Any method you will call on the handle object after
2309destroying it in this way will be silently ignored (and it will return the
2310empty list).
2311
2312Normally, you can just "forget" any references to an AnyEvent::Handle
2313object and it will simply shut down. This works in fatal error and EOF
2314callbacks, as well as code outside. It does I<NOT> work in a read or write
2315callback, so when you want to destroy the AnyEvent::Handle object from
2316within such an callback. You I<MUST> call C<< ->destroy >> explicitly in
2317that case.
2318
2319Destroying the handle object in this way has the advantage that callbacks
2320will be removed as well, so if those are the only reference holders (as
2321is common), then one doesn't need to do anything special to break any
2322reference cycles.
2323
2324The handle might still linger in the background and write out remaining
2325data, as specified by the C<linger> option, however.
2326
2327=cut
2328
2329sub destroy {
2330 my ($self) = @_;
2331
2332 $self->DESTROY;
2333 %$self = ();
2334 bless $self, "AnyEvent::Handle::destroyed";
2335}
2336
2337sub AnyEvent::Handle::destroyed::AUTOLOAD {
2338 #nop
2339}
2340
2341=item $handle->destroyed
2342
2343Returns false as long as the handle hasn't been destroyed by a call to C<<
2344->destroy >>, true otherwise.
2345
2346Can be useful to decide whether the handle is still valid after some
2347callback possibly destroyed the handle. For example, C<< ->push_write >>,
2348C<< ->starttls >> and other methods can call user callbacks, which in turn
2349can destroy the handle, so work can be avoided by checking sometimes:
2350
2351 $hdl->starttls ("accept");
2352 return if $hdl->destroyed;
2353 $hdl->push_write (...
2354
2355Note that the call to C<push_write> will silently be ignored if the handle
2356has been destroyed, so often you can just ignore the possibility of the
2357handle being destroyed.
2358
2359=cut
2360
2361sub destroyed { 0 }
2362sub AnyEvent::Handle::destroyed::destroyed { 1 }
1003 2363
1004=item AnyEvent::Handle::TLS_CTX 2364=item AnyEvent::Handle::TLS_CTX
1005 2365
1006This function creates and returns the Net::SSLeay::CTX object used by 2366This function creates and returns the AnyEvent::TLS object used by default
1007default for TLS mode. 2367for TLS mode.
1008 2368
1009The context is created like this: 2369The context is created by calling L<AnyEvent::TLS> without any arguments.
1010
1011 Net::SSLeay::load_error_strings;
1012 Net::SSLeay::SSLeay_add_ssl_algorithms;
1013 Net::SSLeay::randomize;
1014
1015 my $CTX = Net::SSLeay::CTX_new;
1016
1017 Net::SSLeay::CTX_set_options $CTX, Net::SSLeay::OP_ALL
1018 2370
1019=cut 2371=cut
1020 2372
1021our $TLS_CTX; 2373our $TLS_CTX;
1022 2374
1023sub TLS_CTX() { 2375sub TLS_CTX() {
1024 $TLS_CTX || do { 2376 $TLS_CTX ||= do {
1025 require Net::SSLeay; 2377 require AnyEvent::TLS;
1026 2378
1027 Net::SSLeay::load_error_strings (); 2379 new AnyEvent::TLS
1028 Net::SSLeay::SSLeay_add_ssl_algorithms (); 2380 }
1029 Net::SSLeay::randomize (); 2381}
1030 2382
1031 $TLS_CTX = Net::SSLeay::CTX_new (); 2383=back
1032 2384
1033 Net::SSLeay::CTX_set_options ($TLS_CTX, Net::SSLeay::OP_ALL ());
1034 2385
1035 $TLS_CTX 2386=head1 NONFREQUENTLY ASKED QUESTIONS
2387
2388=over 4
2389
2390=item I C<undef> the AnyEvent::Handle reference inside my callback and
2391still get further invocations!
2392
2393That's because AnyEvent::Handle keeps a reference to itself when handling
2394read or write callbacks.
2395
2396It is only safe to "forget" the reference inside EOF or error callbacks,
2397from within all other callbacks, you need to explicitly call the C<<
2398->destroy >> method.
2399
2400=item Why is my C<on_eof> callback never called?
2401
2402Probably because your C<on_error> callback is being called instead: When
2403you have outstanding requests in your read queue, then an EOF is
2404considered an error as you clearly expected some data.
2405
2406To avoid this, make sure you have an empty read queue whenever your handle
2407is supposed to be "idle" (i.e. connection closes are O.K.). You can set
2408an C<on_read> handler that simply pushes the first read requests in the
2409queue.
2410
2411See also the next question, which explains this in a bit more detail.
2412
2413=item How can I serve requests in a loop?
2414
2415Most protocols consist of some setup phase (authentication for example)
2416followed by a request handling phase, where the server waits for requests
2417and handles them, in a loop.
2418
2419There are two important variants: The first (traditional, better) variant
2420handles requests until the server gets some QUIT command, causing it to
2421close the connection first (highly desirable for a busy TCP server). A
2422client dropping the connection is an error, which means this variant can
2423detect an unexpected detection close.
2424
2425To handle this case, always make sure you have a non-empty read queue, by
2426pushing the "read request start" handler on it:
2427
2428 # we assume a request starts with a single line
2429 my @start_request; @start_request = (line => sub {
2430 my ($hdl, $line) = @_;
2431
2432 ... handle request
2433
2434 # push next request read, possibly from a nested callback
2435 $hdl->push_read (@start_request);
2436 });
2437
2438 # auth done, now go into request handling loop
2439 # now push the first @start_request
2440 $hdl->push_read (@start_request);
2441
2442By always having an outstanding C<push_read>, the handle always expects
2443some data and raises the C<EPIPE> error when the connction is dropped
2444unexpectedly.
2445
2446The second variant is a protocol where the client can drop the connection
2447at any time. For TCP, this means that the server machine may run out of
2448sockets easier, and in general, it means you cannot distinguish a protocl
2449failure/client crash from a normal connection close. Nevertheless, these
2450kinds of protocols are common (and sometimes even the best solution to the
2451problem).
2452
2453Having an outstanding read request at all times is possible if you ignore
2454C<EPIPE> errors, but this doesn't help with when the client drops the
2455connection during a request, which would still be an error.
2456
2457A better solution is to push the initial request read in an C<on_read>
2458callback. This avoids an error, as when the server doesn't expect data
2459(i.e. is idly waiting for the next request, an EOF will not raise an
2460error, but simply result in an C<on_eof> callback. It is also a bit slower
2461and simpler:
2462
2463 # auth done, now go into request handling loop
2464 $hdl->on_read (sub {
2465 my ($hdl) = @_;
2466
2467 # called each time we receive data but the read queue is empty
2468 # simply start read the request
2469
2470 $hdl->push_read (line => sub {
2471 my ($hdl, $line) = @_;
2472
2473 ... handle request
2474
2475 # do nothing special when the request has been handled, just
2476 # let the request queue go empty.
2477 });
2478 });
2479
2480=item I get different callback invocations in TLS mode/Why can't I pause
2481reading?
2482
2483Unlike, say, TCP, TLS connections do not consist of two independent
2484communication channels, one for each direction. Or put differently, the
2485read and write directions are not independent of each other: you cannot
2486write data unless you are also prepared to read, and vice versa.
2487
2488This means that, in TLS mode, you might get C<on_error> or C<on_eof>
2489callback invocations when you are not expecting any read data - the reason
2490is that AnyEvent::Handle always reads in TLS mode.
2491
2492During the connection, you have to make sure that you always have a
2493non-empty read-queue, or an C<on_read> watcher. At the end of the
2494connection (or when you no longer want to use it) you can call the
2495C<destroy> method.
2496
2497=item How do I read data until the other side closes the connection?
2498
2499If you just want to read your data into a perl scalar, the easiest way
2500to achieve this is by setting an C<on_read> callback that does nothing,
2501clearing the C<on_eof> callback and in the C<on_error> callback, the data
2502will be in C<$_[0]{rbuf}>:
2503
2504 $handle->on_read (sub { });
2505 $handle->on_eof (undef);
2506 $handle->on_error (sub {
2507 my $data = delete $_[0]{rbuf};
2508 });
2509
2510Note that this example removes the C<rbuf> member from the handle object,
2511which is not normally allowed by the API. It is expressly permitted in
2512this case only, as the handle object needs to be destroyed afterwards.
2513
2514The reason to use C<on_error> is that TCP connections, due to latencies
2515and packets loss, might get closed quite violently with an error, when in
2516fact all data has been received.
2517
2518It is usually better to use acknowledgements when transferring data,
2519to make sure the other side hasn't just died and you got the data
2520intact. This is also one reason why so many internet protocols have an
2521explicit QUIT command.
2522
2523=item I don't want to destroy the handle too early - how do I wait until
2524all data has been written?
2525
2526After writing your last bits of data, set the C<on_drain> callback
2527and destroy the handle in there - with the default setting of
2528C<low_water_mark> this will be called precisely when all data has been
2529written to the socket:
2530
2531 $handle->push_write (...);
2532 $handle->on_drain (sub {
2533 AE::log debug => "All data submitted to the kernel.";
2534 undef $handle;
2535 });
2536
2537If you just want to queue some data and then signal EOF to the other side,
2538consider using C<< ->push_shutdown >> instead.
2539
2540=item I want to contact a TLS/SSL server, I don't care about security.
2541
2542If your TLS server is a pure TLS server (e.g. HTTPS) that only speaks TLS,
2543connect to it and then create the AnyEvent::Handle with the C<tls>
2544parameter:
2545
2546 tcp_connect $host, $port, sub {
2547 my ($fh) = @_;
2548
2549 my $handle = new AnyEvent::Handle
2550 fh => $fh,
2551 tls => "connect",
2552 on_error => sub { ... };
2553
2554 $handle->push_write (...);
1036 } 2555 };
1037} 2556
2557=item I want to contact a TLS/SSL server, I do care about security.
2558
2559Then you should additionally enable certificate verification, including
2560peername verification, if the protocol you use supports it (see
2561L<AnyEvent::TLS>, C<verify_peername>).
2562
2563E.g. for HTTPS:
2564
2565 tcp_connect $host, $port, sub {
2566 my ($fh) = @_;
2567
2568 my $handle = new AnyEvent::Handle
2569 fh => $fh,
2570 peername => $host,
2571 tls => "connect",
2572 tls_ctx => { verify => 1, verify_peername => "https" },
2573 ...
2574
2575Note that you must specify the hostname you connected to (or whatever
2576"peername" the protocol needs) as the C<peername> argument, otherwise no
2577peername verification will be done.
2578
2579The above will use the system-dependent default set of trusted CA
2580certificates. If you want to check against a specific CA, add the
2581C<ca_file> (or C<ca_cert>) arguments to C<tls_ctx>:
2582
2583 tls_ctx => {
2584 verify => 1,
2585 verify_peername => "https",
2586 ca_file => "my-ca-cert.pem",
2587 },
2588
2589=item I want to create a TLS/SSL server, how do I do that?
2590
2591Well, you first need to get a server certificate and key. You have
2592three options: a) ask a CA (buy one, use cacert.org etc.) b) create a
2593self-signed certificate (cheap. check the search engine of your choice,
2594there are many tutorials on the net) or c) make your own CA (tinyca2 is a
2595nice program for that purpose).
2596
2597Then create a file with your private key (in PEM format, see
2598L<AnyEvent::TLS>), followed by the certificate (also in PEM format). The
2599file should then look like this:
2600
2601 -----BEGIN RSA PRIVATE KEY-----
2602 ...header data
2603 ... lots of base64'y-stuff
2604 -----END RSA PRIVATE KEY-----
2605
2606 -----BEGIN CERTIFICATE-----
2607 ... lots of base64'y-stuff
2608 -----END CERTIFICATE-----
2609
2610The important bits are the "PRIVATE KEY" and "CERTIFICATE" parts. Then
2611specify this file as C<cert_file>:
2612
2613 tcp_server undef, $port, sub {
2614 my ($fh) = @_;
2615
2616 my $handle = new AnyEvent::Handle
2617 fh => $fh,
2618 tls => "accept",
2619 tls_ctx => { cert_file => "my-server-keycert.pem" },
2620 ...
2621
2622When you have intermediate CA certificates that your clients might not
2623know about, just append them to the C<cert_file>.
1038 2624
1039=back 2625=back
1040 2626
1041=head1 SUBCLASSING AnyEvent::Handle 2627=head1 SUBCLASSING AnyEvent::Handle
1042 2628
1048=over 4 2634=over 4
1049 2635
1050=item * all constructor arguments become object members. 2636=item * all constructor arguments become object members.
1051 2637
1052At least initially, when you pass a C<tls>-argument to the constructor it 2638At least initially, when you pass a C<tls>-argument to the constructor it
1053will end up in C<< $handle->{tls} >>. Those members might be changes or 2639will end up in C<< $handle->{tls} >>. Those members might be changed or
1054mutated later on (for example C<tls> will hold the TLS connection object). 2640mutated later on (for example C<tls> will hold the TLS connection object).
1055 2641
1056=item * other object member names are prefixed with an C<_>. 2642=item * other object member names are prefixed with an C<_>.
1057 2643
1058All object members not explicitly documented (internal use) are prefixed 2644All object members not explicitly documented (internal use) are prefixed
1061 2647
1062=item * all members not documented here and not prefixed with an underscore 2648=item * all members not documented here and not prefixed with an underscore
1063are free to use in subclasses. 2649are free to use in subclasses.
1064 2650
1065Of course, new versions of AnyEvent::Handle may introduce more "public" 2651Of course, new versions of AnyEvent::Handle may introduce more "public"
1066member variables, but thats just life, at least it is documented. 2652member variables, but that's just life. At least it is documented.
1067 2653
1068=back 2654=back
1069 2655
1070=head1 AUTHOR 2656=head1 AUTHOR
1071 2657
1072Robin Redeker C<< <elmex at ta-sa.org> >>, Marc Lehmann <schmorp@schmorp.de>. 2658Robin Redeker C<< <elmex at ta-sa.org> >>, Marc Lehmann <schmorp@schmorp.de>.
1073 2659
1074=cut 2660=cut
1075 2661
10761; # End of AnyEvent::Handle 26621
2663

Diff Legend

Removed lines
+ Added lines
< Changed lines
> Changed lines