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 1.61 by root, Fri Jun 6 10:23:50 2008 UTC vs.
Revision 1.243 by root, Mon Mar 16 08:15:46 2015 UTC

1package AnyEvent::Handle;
2
3no warnings;
4use strict;
5
6use AnyEvent ();
7use AnyEvent::Util qw(WSAEWOULDBLOCK);
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 = 4.14;
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 EWOULDBLOCK 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 => $filehandle [C<fh> or C<connect> MANDATORY]
72 93
73The filehandle this L<AnyEvent::Handle> object will operate on. 94The filehandle this L<AnyEvent::Handle> object will operate on.
74
75NOTE: The filehandle will be set to non-blocking (using 95NOTE: The filehandle will be set to non-blocking mode (using
76AnyEvent::Util::fh_nonblocking). 96C<AnyEvent::Util::fh_nonblocking>) by the constructor and needs to stay in
97that mode.
77 98
99=item connect => [$host, $service] [C<fh> or C<connect> MANDATORY]
100
101Try to connect to the specified host and service (port), using
102C<AnyEvent::Socket::tcp_connect>. The C<$host> additionally becomes the
103default C<peername>.
104
105You have to specify either this parameter, or C<fh>, above.
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
78=item on_eof => $cb->($handle) 116=item on_prepare => $cb->($handle)
79 117
80Set the callback to be called when an end-of-file condition is detcted, 118This (rarely used) callback is called before a new connection is
81i.e. in the case of a socket, when the other side has closed the 119attempted, but after the file handle has been created (you can access that
82connection cleanly. 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).
83 123
84While not mandatory, it is highly recommended to set an eof callback, 124The return value of this callback should be the connect timeout value in
85otherwise you might end up with a closed socket while you are still 125seconds (or C<0>, or C<undef>, or the empty list, to indicate that the
86waiting for data. 126default timeout is to be used).
87 127
128=item on_connect => $cb->($handle, $host, $port, $retry->())
129
130This callback is called when a connection has been successfully established.
131
132The peer's numeric host and port (the socket peername) are passed as
133parameters, together with a retry callback. At the time it is called the
134read and write queues, EOF status, TLS status and similar properties of
135the handle will have been reset.
136
137If, for some reason, the handle is not acceptable, calling C<$retry> will
138continue with the next connection target (in case of multi-homed hosts or
139SRV records there can be multiple connection endpoints). The C<$retry>
140callback can be invoked after the connect callback returns, i.e. one can
141start a handshake and then decide to retry with the next host if the
142handshake fails.
143
144In most cases, you should ignore the C<$retry> parameter.
145
146=item on_connect_error => $cb->($handle, $message)
147
148This callback is called when the connection could not be
149established. C<$!> will contain the relevant error code, and C<$message> a
150message describing it (usually the same as C<"$!">).
151
152If this callback isn't specified, then C<on_error> will be called with a
153fatal error instead.
154
155=back
156
88=item on_error => $cb->($handle, $fatal) 157=item on_error => $cb->($handle, $fatal, $message)
89 158
90This is the error callback, which is called when, well, some error 159This is the error callback, which is called when, well, some error
91occured, such as not being able to resolve the hostname, failure to 160occured, such as not being able to resolve the hostname, failure to
92connect or a read error. 161connect, or a read error.
93 162
94Some errors are fatal (which is indicated by C<$fatal> being true). On 163Some errors are fatal (which is indicated by C<$fatal> being true). On
95fatal errors the handle object will be shut down and will not be 164fatal errors the handle object will be destroyed (by a call to C<< ->
165destroy >>) after invoking the error callback (which means you are free to
166examine the handle object). Examples of fatal errors are an EOF condition
167with active (but unsatisfiable) read watchers (C<EPIPE>) or I/O errors. In
168cases where the other side can close the connection at will, it is
169often easiest to not report C<EPIPE> errors in this callback.
170
171AnyEvent::Handle tries to find an appropriate error code for you to check
172against, but in some cases (TLS errors), this does not work well.
173
174If you report the error to the user, it is recommended to always output
175the C<$message> argument in human-readable error messages (you don't need
176to report C<"$!"> if you report C<$message>).
177
178If you want to react programmatically to the error, then looking at C<$!>
179and comparing it against some of the documented C<Errno> values is usually
180better than looking at the C<$message>.
181
96usable. Non-fatal errors can be retried by simply returning, but it is 182Non-fatal errors can be retried by returning, but it is recommended
97recommended to simply ignore this parameter and instead abondon the handle 183to simply ignore this parameter and instead abondon the handle object
98object when this callback is invoked. 184when this callback is invoked. Examples of non-fatal errors are timeouts
185C<ETIMEDOUT>) or badly-formatted data (C<EBADMSG>).
99 186
100On callback entrance, the value of C<$!> contains the operating system 187On entry to the callback, the value of C<$!> contains the operating
101error (or C<ENOSPC>, C<EPIPE>, C<ETIMEDOUT> or C<EBADMSG>). 188system error code (or C<ENOSPC>, C<EPIPE>, C<ETIMEDOUT>, C<EBADMSG> or
189C<EPROTO>).
102 190
103While not mandatory, it is I<highly> recommended to set this callback, as 191While not mandatory, it is I<highly> recommended to set this callback, as
104you will not be notified of errors otherwise. The default simply calls 192you will not be notified of errors otherwise. The default just calls
105C<croak>. 193C<croak>.
106 194
107=item on_read => $cb->($handle) 195=item on_read => $cb->($handle)
108 196
109This sets the default read callback, which is called when data arrives 197This sets the default read callback, which is called when data arrives
110and no read request is in the queue (unlike read queue callbacks, this 198and no read request is in the queue (unlike read queue callbacks, this
111callback will only be called when at least one octet of data is in the 199callback will only be called when at least one octet of data is in the
112read buffer). 200read buffer).
113 201
114To access (and remove data from) the read buffer, use the C<< ->rbuf >> 202To access (and remove data from) the read buffer, use the C<< ->rbuf >>
115method or access the C<$handle->{rbuf}> member directly. 203method or access the C<< $handle->{rbuf} >> member directly. Note that you
204must not enlarge or modify the read buffer, you can only remove data at
205the beginning from it.
116 206
207You can also call C<< ->push_read (...) >> or any other function that
208modifies the read queue. Or do both. Or ...
209
117When an EOF condition is detected then AnyEvent::Handle will first try to 210When an EOF condition is detected, AnyEvent::Handle will first try to
118feed all the remaining data to the queued callbacks and C<on_read> before 211feed all the remaining data to the queued callbacks and C<on_read> before
119calling the C<on_eof> callback. If no progress can be made, then a fatal 212calling the C<on_eof> callback. If no progress can be made, then a fatal
120error will be raised (with C<$!> set to C<EPIPE>). 213error will be raised (with C<$!> set to C<EPIPE>).
121 214
215Note that, unlike requests in the read queue, an C<on_read> callback
216doesn't mean you I<require> some data: if there is an EOF and there
217are outstanding read requests then an error will be flagged. With an
218C<on_read> callback, the C<on_eof> callback will be invoked.
219
220=item on_eof => $cb->($handle)
221
222Set the callback to be called when an end-of-file condition is detected,
223i.e. in the case of a socket, when the other side has closed the
224connection cleanly, and there are no outstanding read requests in the
225queue (if there are read requests, then an EOF counts as an unexpected
226connection close and will be flagged as an error).
227
228For sockets, this just means that the other side has stopped sending data,
229you can still try to write data, and, in fact, one can return from the EOF
230callback and continue writing data, as only the read part has been shut
231down.
232
233If an EOF condition has been detected but no C<on_eof> callback has been
234set, then a fatal error will be raised with C<$!> set to <0>.
235
122=item on_drain => $cb->($handle) 236=item on_drain => $cb->($handle)
123 237
124This sets the callback that is called when the write buffer becomes empty 238This sets the callback that is called once when the write buffer becomes
125(or when the callback is set and the buffer is empty already). 239empty (and immediately when the handle object is created).
126 240
127To append to the write buffer, use the C<< ->push_write >> method. 241To append to the write buffer, use the C<< ->push_write >> method.
128 242
243This callback is useful when you don't want to put all of your write data
244into the queue at once, for example, when you want to write the contents
245of some file to the socket you might not want to read the whole file into
246memory and push it into the queue, but instead only read more data from
247the file when the write queue becomes empty.
248
129=item timeout => $fractional_seconds 249=item timeout => $fractional_seconds
130 250
251=item rtimeout => $fractional_seconds
252
253=item wtimeout => $fractional_seconds
254
131If non-zero, then this enables an "inactivity" timeout: whenever this many 255If non-zero, then these enables an "inactivity" timeout: whenever this
132seconds pass without a successful read or write on the underlying file 256many seconds pass without a successful read or write on the underlying
133handle, the C<on_timeout> callback will be invoked (and if that one is 257file handle (or a call to C<timeout_reset>), the C<on_timeout> callback
134missing, an C<ETIMEDOUT> error will be raised). 258will be invoked (and if that one is missing, a non-fatal C<ETIMEDOUT>
259error will be raised).
135 260
261There are three variants of the timeouts that work independently of each
262other, for both read and write (triggered when nothing was read I<OR>
263written), just read (triggered when nothing was read), and just write:
264C<timeout>, C<rtimeout> and C<wtimeout>, with corresponding callbacks
265C<on_timeout>, C<on_rtimeout> and C<on_wtimeout>, and reset functions
266C<timeout_reset>, C<rtimeout_reset>, and C<wtimeout_reset>.
267
136Note that timeout processing is also active when you currently do not have 268Note that timeout processing is active even when you do not have any
137any outstanding read or write requests: If you plan to keep the connection 269outstanding read or write requests: If you plan to keep the connection
138idle then you should disable the timout temporarily or ignore the timeout 270idle then you should disable the timeout temporarily or ignore the
139in the C<on_timeout> callback. 271timeout in the corresponding C<on_timeout> callback, in which case
272AnyEvent::Handle will simply restart the timeout.
140 273
141Zero (the default) disables this timeout. 274Zero (the default) disables the corresponding timeout.
142 275
143=item on_timeout => $cb->($handle) 276=item on_timeout => $cb->($handle)
277
278=item on_rtimeout => $cb->($handle)
279
280=item on_wtimeout => $cb->($handle)
144 281
145Called whenever the inactivity timeout passes. If you return from this 282Called whenever the inactivity timeout passes. If you return from this
146callback, then the timeout will be reset as if some activity had happened, 283callback, then the timeout will be reset as if some activity had happened,
147so this condition is not fatal in any way. 284so this condition is not fatal in any way.
148 285
149=item rbuf_max => <bytes> 286=item rbuf_max => <bytes>
150 287
151If defined, then a fatal error will be raised (with C<$!> set to C<ENOSPC>) 288If defined, then a fatal error will be raised (with C<$!> set to C<ENOSPC>)
152when the read buffer ever (strictly) exceeds this size. This is useful to 289when the read buffer ever (strictly) exceeds this size. This is useful to
153avoid denial-of-service attacks. 290avoid some forms of denial-of-service attacks.
154 291
155For example, a server accepting connections from untrusted sources should 292For example, a server accepting connections from untrusted sources should
156be configured to accept only so-and-so much data that it cannot act on 293be configured to accept only so-and-so much data that it cannot act on
157(for example, when expecting a line, an attacker could send an unlimited 294(for example, when expecting a line, an attacker could send an unlimited
158amount of data without a callback ever being called as long as the line 295amount of data without a callback ever being called as long as the line
159isn't finished). 296isn't finished).
160 297
298=item wbuf_max => <bytes>
299
300If defined, then a fatal error will be raised (with C<$!> set to C<ENOSPC>)
301when the write buffer ever (strictly) exceeds this size. This is useful to
302avoid some forms of denial-of-service attacks.
303
304Although the units of this parameter is bytes, this is the I<raw> number
305of bytes not yet accepted by the kernel. This can make a difference when
306you e.g. use TLS, as TLS typically makes your write data larger (but it
307can also make it smaller due to compression).
308
309As an example of when this limit is useful, take a chat server that sends
310chat messages to a client. If the client does not read those in a timely
311manner then the send buffer in the server would grow unbounded.
312
313=item autocork => <boolean>
314
315When disabled (the default), C<push_write> will try to immediately
316write the data to the handle if possible. This avoids having to register
317a write watcher and wait for the next event loop iteration, but can
318be inefficient if you write multiple small chunks (on the wire, this
319disadvantage is usually avoided by your kernel's nagle algorithm, see
320C<no_delay>, but this option can save costly syscalls).
321
322When enabled, writes will always be queued till the next event loop
323iteration. This is efficient when you do many small writes per iteration,
324but less efficient when you do a single write only per iteration (or when
325the write buffer often is full). It also increases write latency.
326
327=item no_delay => <boolean>
328
329When doing small writes on sockets, your operating system kernel might
330wait a bit for more data before actually sending it out. This is called
331the Nagle algorithm, and usually it is beneficial.
332
333In some situations you want as low a delay as possible, which can be
334accomplishd by setting this option to a true value.
335
336The default is your operating system's default behaviour (most likely
337enabled). This option explicitly enables or disables it, if possible.
338
339=item keepalive => <boolean>
340
341Enables (default disable) the SO_KEEPALIVE option on the stream socket:
342normally, TCP connections have no time-out once established, so TCP
343connections, once established, can stay alive forever even when the other
344side has long gone. TCP keepalives are a cheap way to take down long-lived
345TCP connections when the other side becomes unreachable. While the default
346is OS-dependent, TCP keepalives usually kick in after around two hours,
347and, if the other side doesn't reply, take down the TCP connection some 10
348to 15 minutes later.
349
350It is harmless to specify this option for file handles that do not support
351keepalives, and enabling it on connections that are potentially long-lived
352is usually a good idea.
353
354=item oobinline => <boolean>
355
356BSD majorly fucked up the implementation of TCP urgent data. The result
357is that almost no OS implements TCP according to the specs, and every OS
358implements it slightly differently.
359
360If you want to handle TCP urgent data, then setting this flag (the default
361is enabled) gives you the most portable way of getting urgent data, by
362putting it into the stream.
363
364Since BSD emulation of OOB data on top of TCP's urgent data can have
365security implications, AnyEvent::Handle sets this flag automatically
366unless explicitly specified. Note that setting this flag after
367establishing a connection I<may> be a bit too late (data loss could
368already have occured on BSD systems), but at least it will protect you
369from most attacks.
370
161=item read_size => <bytes> 371=item read_size => <bytes>
162 372
163The default read block size (the amount of bytes this module will try to read 373The initial read block size, the number of bytes this module will try
164during each (loop iteration). Default: C<8192>. 374to read during each loop iteration. Each handle object will consume
375at least this amount of memory for the read buffer as well, so when
376handling many connections watch out for memory requirements). See also
377C<max_read_size>. Default: C<2048>.
378
379=item max_read_size => <bytes>
380
381The maximum read buffer size used by the dynamic adjustment
382algorithm: Each time AnyEvent::Handle can read C<read_size> bytes in
383one go it will double C<read_size> up to the maximum given by this
384option. Default: C<131072> or C<read_size>, whichever is higher.
165 385
166=item low_water_mark => <bytes> 386=item low_water_mark => <bytes>
167 387
168Sets the amount of bytes (default: C<0>) that make up an "empty" write 388Sets the number of bytes (default: C<0>) that make up an "empty" write
169buffer: If the write reaches this size or gets even samller it is 389buffer: If the buffer reaches this size or gets even samller it is
170considered empty. 390considered empty.
171 391
392Sometimes it can be beneficial (for performance reasons) to add data to
393the write buffer before it is fully drained, but this is a rare case, as
394the operating system kernel usually buffers data as well, so the default
395is good in almost all cases.
396
397=item linger => <seconds>
398
399If this is non-zero (default: C<3600>), the destructor of the
400AnyEvent::Handle object will check whether there is still outstanding
401write data and will install a watcher that will write this data to the
402socket. No errors will be reported (this mostly matches how the operating
403system treats outstanding data at socket close time).
404
405This will not work for partial TLS data that could not be encoded
406yet. This data will be lost. Calling the C<stoptls> method in time might
407help.
408
409=item peername => $string
410
411A string used to identify the remote site - usually the DNS hostname
412(I<not> IDN!) used to create the connection, rarely the IP address.
413
414Apart from being useful in error messages, this string is also used in TLS
415peername verification (see C<verify_peername> in L<AnyEvent::TLS>). This
416verification will be skipped when C<peername> is not specified or is
417C<undef>.
418
172=item tls => "accept" | "connect" | Net::SSLeay::SSL object 419=item tls => "accept" | "connect" | Net::SSLeay::SSL object
173 420
174When this parameter is given, it enables TLS (SSL) mode, that means it 421When this parameter is given, it enables TLS (SSL) mode, that means
175will start making tls handshake and will transparently encrypt/decrypt 422AnyEvent will start a TLS handshake as soon as the connection has been
176data. 423established and will transparently encrypt/decrypt data afterwards.
424
425All TLS protocol errors will be signalled as C<EPROTO>, with an
426appropriate error message.
177 427
178TLS mode requires Net::SSLeay to be installed (it will be loaded 428TLS mode requires Net::SSLeay to be installed (it will be loaded
179automatically when you try to create a TLS handle). 429automatically when you try to create a TLS handle): this module doesn't
430have a dependency on that module, so if your module requires it, you have
431to add the dependency yourself. If Net::SSLeay cannot be loaded or is too
432old, you get an C<EPROTO> error.
180 433
181For the TLS server side, use C<accept>, and for the TLS client side of a 434Unlike TCP, TLS has a server and client side: for the TLS server side, use
182connection, use C<connect> mode. 435C<accept>, and for the TLS client side of a connection, use C<connect>
436mode.
183 437
184You can also provide your own TLS connection object, but you have 438You can also provide your own TLS connection object, but you have
185to make sure that you call either C<Net::SSLeay::set_connect_state> 439to make sure that you call either C<Net::SSLeay::set_connect_state>
186or C<Net::SSLeay::set_accept_state> on it before you pass it to 440or C<Net::SSLeay::set_accept_state> on it before you pass it to
187AnyEvent::Handle. 441AnyEvent::Handle. Also, this module will take ownership of this connection
442object.
188 443
444At some future point, AnyEvent::Handle might switch to another TLS
445implementation, then the option to use your own session object will go
446away.
447
448B<IMPORTANT:> since Net::SSLeay "objects" are really only integers,
449passing in the wrong integer will lead to certain crash. This most often
450happens when one uses a stylish C<< tls => 1 >> and is surprised about the
451segmentation fault.
452
189See the C<starttls> method if you need to start TLs negotiation later. 453Use the C<< ->starttls >> method if you need to start TLS negotiation later.
190 454
191=item tls_ctx => $ssl_ctx 455=item tls_ctx => $anyevent_tls
192 456
193Use the given Net::SSLeay::CTX object to create the new TLS connection 457Use the given C<AnyEvent::TLS> object to create the new TLS connection
194(unless a connection object was specified directly). If this parameter is 458(unless a connection object was specified directly). If this
195missing, then AnyEvent::Handle will use C<AnyEvent::Handle::TLS_CTX>. 459parameter is missing (or C<undef>), then AnyEvent::Handle will use
460C<AnyEvent::Handle::TLS_CTX>.
196 461
462Instead of an object, you can also specify a hash reference with C<< key
463=> value >> pairs. Those will be passed to L<AnyEvent::TLS> to create a
464new TLS context object.
465
466=item on_starttls => $cb->($handle, $success[, $error_message])
467
468This callback will be invoked when the TLS/SSL handshake has finished. If
469C<$success> is true, then the TLS handshake succeeded, otherwise it failed
470(C<on_stoptls> will not be called in this case).
471
472The session in C<< $handle->{tls} >> can still be examined in this
473callback, even when the handshake was not successful.
474
475TLS handshake failures will not cause C<on_error> to be invoked when this
476callback is in effect, instead, the error message will be passed to C<on_starttls>.
477
478Without this callback, handshake failures lead to C<on_error> being
479called as usual.
480
481Note that you cannot just call C<starttls> again in this callback. If you
482need to do that, start an zero-second timer instead whose callback can
483then call C<< ->starttls >> again.
484
485=item on_stoptls => $cb->($handle)
486
487When a SSLv3/TLS shutdown/close notify/EOF is detected and this callback is
488set, then it will be invoked after freeing the TLS session. If it is not,
489then a TLS shutdown condition will be treated like a normal EOF condition
490on the handle.
491
492The session in C<< $handle->{tls} >> can still be examined in this
493callback.
494
495This callback will only be called on TLS shutdowns, not when the
496underlying handle signals EOF.
497
197=item json => JSON or JSON::XS object 498=item json => L<JSON>, L<JSON::PP> or L<JSON::XS> object
198 499
199This is the json coder object used by the C<json> read and write types. 500This is the json coder object used by the C<json> read and write types.
200 501
201If you don't supply it, then AnyEvent::Handle will create and use a 502If you don't supply it, then AnyEvent::Handle will create and use a
202suitable one, which will write and expect UTF-8 encoded JSON texts. 503suitable one (on demand), which will write and expect UTF-8 encoded
504JSON texts (either using L<JSON::XS> or L<JSON>). The written texts are
505guaranteed not to contain any newline character.
203 506
507For security reasons, this encoder will likely I<not> handle numbers and
508strings, only arrays and objects/hashes. The reason is that originally
509JSON was self-delimited, but Dougles Crockford thought it was a splendid
510idea to redefine JSON incompatibly, so this is no longer true.
511
512For protocols that used back-to-back JSON texts, this might lead to
513run-ins, where two or more JSON texts will be interpreted as one JSON
514text.
515
516For this reason, if the default encoder uses L<JSON::XS>, it will default
517to not allowing anything but arrays and objects/hashes, at least for the
518forseeable future (it will change at some point). This might or might not
519be true for the L<JSON> module, so this might cause a security issue.
520
521If you depend on either behaviour, you should create your own json object
522and pass it in explicitly.
523
524=item cbor => L<CBOR::XS> object
525
526This is the cbor coder object used by the C<cbor> read and write types.
527
528If you don't supply it, then AnyEvent::Handle will create and use a
529suitable one (on demand), which will write CBOR without using extensions,
530if possible.
531
204Note that you are responsible to depend on the JSON module if you want to 532Note that you are responsible to depend on the L<CBOR::XS> module if you
205use this functionality, as AnyEvent does not have a dependency itself. 533want to use this functionality, as AnyEvent does not have a dependency on
206 534it itself.
207=item filter_r => $cb
208
209=item filter_w => $cb
210
211These exist, but are undocumented at this time.
212 535
213=back 536=back
214 537
215=cut 538=cut
216 539
217sub new { 540sub new {
218 my $class = shift; 541 my $class = shift;
219
220 my $self = bless { @_ }, $class; 542 my $self = bless { @_ }, $class;
221 543
222 $self->{fh} or Carp::croak "mandatory argument fh is missing"; 544 if ($self->{fh}) {
545 $self->_start;
546 return unless $self->{fh}; # could be gone by now
547
548 } elsif ($self->{connect}) {
549 require AnyEvent::Socket;
550
551 $self->{peername} = $self->{connect}[0]
552 unless exists $self->{peername};
553
554 $self->{_skip_drain_rbuf} = 1;
555
556 {
557 Scalar::Util::weaken (my $self = $self);
558
559 $self->{_connect} =
560 AnyEvent::Socket::tcp_connect (
561 $self->{connect}[0],
562 $self->{connect}[1],
563 sub {
564 my ($fh, $host, $port, $retry) = @_;
565
566 delete $self->{_connect}; # no longer needed
567
568 if ($fh) {
569 $self->{fh} = $fh;
570
571 delete $self->{_skip_drain_rbuf};
572 $self->_start;
573
574 $self->{on_connect}
575 and $self->{on_connect}($self, $host, $port, sub {
576 delete @$self{qw(fh _tw _rtw _wtw _ww _rw _eof _queue rbuf _wbuf tls _tls_rbuf _tls_wbuf)};
577 $self->{_skip_drain_rbuf} = 1;
578 &$retry;
579 });
580
581 } else {
582 if ($self->{on_connect_error}) {
583 $self->{on_connect_error}($self, "$!");
584 $self->destroy if $self;
585 } else {
586 $self->_error ($!, 1);
587 }
588 }
589 },
590 sub {
591 local $self->{fh} = $_[0];
592
593 $self->{on_prepare}
594 ? $self->{on_prepare}->($self)
595 : ()
596 }
597 );
598 }
599
600 } else {
601 Carp::croak "AnyEvent::Handle: either an existing fh or the connect parameter must be specified";
602 }
603
604 $self
605}
606
607sub _start {
608 my ($self) = @_;
609
610 # too many clueless people try to use udp and similar sockets
611 # with AnyEvent::Handle, do them a favour.
612 my $type = getsockopt $self->{fh}, Socket::SOL_SOCKET (), Socket::SO_TYPE ();
613 Carp::croak "AnyEvent::Handle: only stream sockets supported, anything else will NOT work!"
614 if Socket::SOCK_STREAM () != (unpack "I", $type) && defined $type;
223 615
224 AnyEvent::Util::fh_nonblocking $self->{fh}, 1; 616 AnyEvent::Util::fh_nonblocking $self->{fh}, 1;
225 617
226 if ($self->{tls}) { 618 $self->{_activity} =
227 require Net::SSLeay; 619 $self->{_ractivity} =
620 $self->{_wactivity} = AE::now;
621
622 $self->{read_size} ||= 2048;
623 $self->{max_read_size} = $self->{read_size}
624 if $self->{read_size} > ($self->{max_read_size} || MAX_READ_SIZE);
625
626 $self->timeout (delete $self->{timeout} ) if $self->{timeout};
627 $self->rtimeout (delete $self->{rtimeout} ) if $self->{rtimeout};
628 $self->wtimeout (delete $self->{wtimeout} ) if $self->{wtimeout};
629
630 $self->no_delay (delete $self->{no_delay} ) if exists $self->{no_delay} && $self->{no_delay};
631 $self->keepalive (delete $self->{keepalive}) if exists $self->{keepalive} && $self->{keepalive};
632
633 $self->oobinline (exists $self->{oobinline} ? delete $self->{oobinline} : 1);
634
228 $self->starttls (delete $self->{tls}, delete $self->{tls_ctx}); 635 $self->starttls (delete $self->{tls}, delete $self->{tls_ctx})
229 } 636 if $self->{tls};
230 637
231 $self->{_activity} = AnyEvent->now;
232 $self->_timeout;
233
234 $self->on_drain (delete $self->{on_drain}) if $self->{on_drain}; 638 $self->on_drain (delete $self->{on_drain} ) if $self->{on_drain};
235 639
236 $self 640 $self->start_read
237} 641 if $self->{on_read} || @{ $self->{_queue} };
238 642
239sub _shutdown { 643 $self->_drain_wbuf;
240 my ($self) = @_;
241
242 delete $self->{_tw};
243 delete $self->{_rw};
244 delete $self->{_ww};
245 delete $self->{fh};
246
247 $self->stoptls;
248} 644}
249 645
250sub _error { 646sub _error {
251 my ($self, $errno, $fatal) = @_; 647 my ($self, $errno, $fatal, $message) = @_;
252
253 $self->_shutdown
254 if $fatal;
255 648
256 $! = $errno; 649 $! = $errno;
650 $message ||= "$!";
257 651
258 if ($self->{on_error}) { 652 if ($self->{on_error}) {
259 $self->{on_error}($self, $fatal); 653 $self->{on_error}($self, $fatal, $message);
260 } else { 654 $self->destroy if $fatal;
655 } elsif ($self->{fh} || $self->{connect}) {
656 $self->destroy;
261 Carp::croak "AnyEvent::Handle uncaught error: $!"; 657 Carp::croak "AnyEvent::Handle uncaught error: $message";
262 } 658 }
263} 659}
264 660
265=item $fh = $handle->fh 661=item $fh = $handle->fh
266 662
267This method returns the file handle of the L<AnyEvent::Handle> object. 663This method returns the file handle used to create the L<AnyEvent::Handle> object.
268 664
269=cut 665=cut
270 666
271sub fh { $_[0]{fh} } 667sub fh { $_[0]{fh} }
272 668
290 $_[0]{on_eof} = $_[1]; 686 $_[0]{on_eof} = $_[1];
291} 687}
292 688
293=item $handle->on_timeout ($cb) 689=item $handle->on_timeout ($cb)
294 690
295Replace the current C<on_timeout> callback, or disables the callback 691=item $handle->on_rtimeout ($cb)
296(but not the timeout) if C<$cb> = C<undef>. See C<timeout> constructor
297argument.
298 692
299=cut 693=item $handle->on_wtimeout ($cb)
300 694
695Replace the current C<on_timeout>, C<on_rtimeout> or C<on_wtimeout>
696callback, or disables the callback (but not the timeout) if C<$cb> =
697C<undef>. See the C<timeout> constructor argument and method.
698
699=cut
700
701# see below
702
703=item $handle->autocork ($boolean)
704
705Enables or disables the current autocork behaviour (see C<autocork>
706constructor argument). Changes will only take effect on the next write.
707
708=cut
709
710sub autocork {
711 $_[0]{autocork} = $_[1];
712}
713
714=item $handle->no_delay ($boolean)
715
716Enables or disables the C<no_delay> setting (see constructor argument of
717the same name for details).
718
719=cut
720
721sub no_delay {
722 $_[0]{no_delay} = $_[1];
723
724 setsockopt $_[0]{fh}, Socket::IPPROTO_TCP (), Socket::TCP_NODELAY (), int $_[1]
725 if $_[0]{fh};
726}
727
728=item $handle->keepalive ($boolean)
729
730Enables or disables the C<keepalive> setting (see constructor argument of
731the same name for details).
732
733=cut
734
735sub keepalive {
736 $_[0]{keepalive} = $_[1];
737
738 eval {
739 local $SIG{__DIE__};
740 setsockopt $_[0]{fh}, Socket::SOL_SOCKET (), Socket::SO_KEEPALIVE (), int $_[1]
741 if $_[0]{fh};
742 };
743}
744
745=item $handle->oobinline ($boolean)
746
747Enables or disables the C<oobinline> setting (see constructor argument of
748the same name for details).
749
750=cut
751
752sub oobinline {
753 $_[0]{oobinline} = $_[1];
754
755 eval {
756 local $SIG{__DIE__};
757 setsockopt $_[0]{fh}, Socket::SOL_SOCKET (), Socket::SO_OOBINLINE (), int $_[1]
758 if $_[0]{fh};
759 };
760}
761
762=item $handle->keepalive ($boolean)
763
764Enables or disables the C<keepalive> setting (see constructor argument of
765the same name for details).
766
767=cut
768
769sub keepalive {
770 $_[0]{keepalive} = $_[1];
771
772 eval {
773 local $SIG{__DIE__};
774 setsockopt $_[0]{fh}, Socket::SOL_SOCKET (), Socket::SO_KEEPALIVE (), int $_[1]
775 if $_[0]{fh};
776 };
777}
778
779=item $handle->on_starttls ($cb)
780
781Replace the current C<on_starttls> callback (see the C<on_starttls> constructor argument).
782
783=cut
784
785sub on_starttls {
786 $_[0]{on_starttls} = $_[1];
787}
788
789=item $handle->on_stoptls ($cb)
790
791Replace the current C<on_stoptls> callback (see the C<on_stoptls> constructor argument).
792
793=cut
794
301sub on_timeout { 795sub on_stoptls {
302 $_[0]{on_timeout} = $_[1]; 796 $_[0]{on_stoptls} = $_[1];
797}
798
799=item $handle->rbuf_max ($max_octets)
800
801Configures the C<rbuf_max> setting (C<undef> disables it).
802
803=item $handle->wbuf_max ($max_octets)
804
805Configures the C<wbuf_max> setting (C<undef> disables it).
806
807=cut
808
809sub rbuf_max {
810 $_[0]{rbuf_max} = $_[1];
811}
812
813sub wbuf_max {
814 $_[0]{wbuf_max} = $_[1];
303} 815}
304 816
305############################################################################# 817#############################################################################
306 818
307=item $handle->timeout ($seconds) 819=item $handle->timeout ($seconds)
308 820
821=item $handle->rtimeout ($seconds)
822
823=item $handle->wtimeout ($seconds)
824
309Configures (or disables) the inactivity timeout. 825Configures (or disables) the inactivity timeout.
310 826
311=cut 827The timeout will be checked instantly, so this method might destroy the
828handle before it returns.
312 829
313sub timeout { 830=item $handle->timeout_reset
831
832=item $handle->rtimeout_reset
833
834=item $handle->wtimeout_reset
835
836Reset the activity timeout, as if data was received or sent.
837
838These methods are cheap to call.
839
840=cut
841
842for my $dir ("", "r", "w") {
843 my $timeout = "${dir}timeout";
844 my $tw = "_${dir}tw";
845 my $on_timeout = "on_${dir}timeout";
846 my $activity = "_${dir}activity";
847 my $cb;
848
849 *$on_timeout = sub {
850 $_[0]{$on_timeout} = $_[1];
851 };
852
853 *$timeout = sub {
314 my ($self, $timeout) = @_; 854 my ($self, $new_value) = @_;
315 855
856 $new_value >= 0
857 or Carp::croak "AnyEvent::Handle->$timeout called with negative timeout ($new_value), caught";
858
316 $self->{timeout} = $timeout; 859 $self->{$timeout} = $new_value;
317 $self->_timeout; 860 delete $self->{$tw}; &$cb;
318} 861 };
319 862
863 *{"${dir}timeout_reset"} = sub {
864 $_[0]{$activity} = AE::now;
865 };
866
867 # main workhorse:
320# reset the timeout watcher, as neccessary 868 # reset the timeout watcher, as neccessary
321# also check for time-outs 869 # also check for time-outs
322sub _timeout { 870 $cb = sub {
323 my ($self) = @_; 871 my ($self) = @_;
324 872
325 if ($self->{timeout}) { 873 if ($self->{$timeout} && $self->{fh}) {
326 my $NOW = AnyEvent->now; 874 my $NOW = AE::now;
327 875
328 # when would the timeout trigger? 876 # when would the timeout trigger?
329 my $after = $self->{_activity} + $self->{timeout} - $NOW; 877 my $after = $self->{$activity} + $self->{$timeout} - $NOW;
330 878
331 # now or in the past already? 879 # now or in the past already?
332 if ($after <= 0) { 880 if ($after <= 0) {
333 $self->{_activity} = $NOW; 881 $self->{$activity} = $NOW;
334 882
335 if ($self->{on_timeout}) { 883 if ($self->{$on_timeout}) {
336 $self->{on_timeout}($self); 884 $self->{$on_timeout}($self);
337 } else { 885 } else {
338 $self->_error (&Errno::ETIMEDOUT); 886 $self->_error (Errno::ETIMEDOUT);
887 }
888
889 # callback could have changed timeout value, optimise
890 return unless $self->{$timeout};
891
892 # calculate new after
893 $after = $self->{$timeout};
339 } 894 }
340 895
341 # callback could have changed timeout value, optimise 896 Scalar::Util::weaken $self;
342 return unless $self->{timeout}; 897 return unless $self; # ->error could have destroyed $self
343 898
344 # calculate new after 899 $self->{$tw} ||= AE::timer $after, 0, sub {
345 $after = $self->{timeout}; 900 delete $self->{$tw};
901 $cb->($self);
902 };
903 } else {
904 delete $self->{$tw};
346 } 905 }
347
348 Scalar::Util::weaken $self;
349 return unless $self; # ->error could have destroyed $self
350
351 $self->{_tw} ||= AnyEvent->timer (after => $after, cb => sub {
352 delete $self->{_tw};
353 $self->_timeout;
354 });
355 } else {
356 delete $self->{_tw};
357 } 906 }
358} 907}
359 908
360############################################################################# 909#############################################################################
361 910
368 917
369The write queue is very simple: you can add data to its end, and 918The write queue is very simple: you can add data to its end, and
370AnyEvent::Handle will automatically try to get rid of it for you. 919AnyEvent::Handle will automatically try to get rid of it for you.
371 920
372When data could be written and the write buffer is shorter then the low 921When data could be written and the write buffer is shorter then the low
373water mark, the C<on_drain> callback will be invoked. 922water mark, the C<on_drain> callback will be invoked once.
374 923
375=over 4 924=over 4
376 925
377=item $handle->on_drain ($cb) 926=item $handle->on_drain ($cb)
378 927
379Sets the C<on_drain> callback or clears it (see the description of 928Sets the C<on_drain> callback or clears it (see the description of
380C<on_drain> in the constructor). 929C<on_drain> in the constructor).
381 930
931This method may invoke callbacks (and therefore the handle might be
932destroyed after it returns).
933
382=cut 934=cut
383 935
384sub on_drain { 936sub on_drain {
385 my ($self, $cb) = @_; 937 my ($self, $cb) = @_;
386 938
387 $self->{on_drain} = $cb; 939 $self->{on_drain} = $cb;
388 940
389 $cb->($self) 941 $cb->($self)
390 if $cb && $self->{low_water_mark} >= length $self->{wbuf}; 942 if $cb && $self->{low_water_mark} >= (length $self->{wbuf}) + (length $self->{_tls_wbuf});
391} 943}
392 944
393=item $handle->push_write ($data) 945=item $handle->push_write ($data)
394 946
395Queues the given scalar to be written. You can push as much data as you 947Queues the given scalar to be written. You can push as much data as
396want (only limited by the available memory), as C<AnyEvent::Handle> 948you want (only limited by the available memory and C<wbuf_max>), as
397buffers it independently of the kernel. 949C<AnyEvent::Handle> buffers it independently of the kernel.
950
951This method may invoke callbacks (and therefore the handle might be
952destroyed after it returns).
398 953
399=cut 954=cut
400 955
401sub _drain_wbuf { 956sub _drain_wbuf {
402 my ($self) = @_; 957 my ($self) = @_;
406 Scalar::Util::weaken $self; 961 Scalar::Util::weaken $self;
407 962
408 my $cb = sub { 963 my $cb = sub {
409 my $len = syswrite $self->{fh}, $self->{wbuf}; 964 my $len = syswrite $self->{fh}, $self->{wbuf};
410 965
411 if ($len >= 0) { 966 if (defined $len) {
412 substr $self->{wbuf}, 0, $len, ""; 967 substr $self->{wbuf}, 0, $len, "";
413 968
414 $self->{_activity} = AnyEvent->now; 969 $self->{_activity} = $self->{_wactivity} = AE::now;
415 970
416 $self->{on_drain}($self) 971 $self->{on_drain}($self)
417 if $self->{low_water_mark} >= length $self->{wbuf} 972 if $self->{low_water_mark} >= (length $self->{wbuf}) + (length $self->{_tls_wbuf})
418 && $self->{on_drain}; 973 && $self->{on_drain};
419 974
420 delete $self->{_ww} unless length $self->{wbuf}; 975 delete $self->{_ww} unless length $self->{wbuf};
421 } elsif ($! != EAGAIN && $! != EINTR && $! != WSAEWOULDBLOCK) { 976 } elsif ($! != EAGAIN && $! != EINTR && $! != EWOULDBLOCK && $! != WSAEWOULDBLOCK) {
422 $self->_error ($!, 1); 977 $self->_error ($!, 1);
423 } 978 }
424 }; 979 };
425 980
426 # try to write data immediately 981 # try to write data immediately
427 $cb->(); 982 $cb->() unless $self->{autocork};
428 983
429 # if still data left in wbuf, we need to poll 984 # if still data left in wbuf, we need to poll
430 $self->{_ww} = AnyEvent->io (fh => $self->{fh}, poll => "w", cb => $cb) 985 $self->{_ww} = AE::io $self->{fh}, 1, $cb
431 if length $self->{wbuf}; 986 if length $self->{wbuf};
987
988 if (
989 defined $self->{wbuf_max}
990 && $self->{wbuf_max} < length $self->{wbuf}
991 ) {
992 $self->_error (Errno::ENOSPC, 1), return;
993 }
432 }; 994 };
433} 995}
434 996
435our %WH; 997our %WH;
436 998
999# deprecated
437sub register_write_type($$) { 1000sub register_write_type($$) {
438 $WH{$_[0]} = $_[1]; 1001 $WH{$_[0]} = $_[1];
439} 1002}
440 1003
441sub push_write { 1004sub push_write {
442 my $self = shift; 1005 my $self = shift;
443 1006
444 if (@_ > 1) { 1007 if (@_ > 1) {
445 my $type = shift; 1008 my $type = shift;
446 1009
1010 @_ = ($WH{$type} ||= _load_func "$type\::anyevent_write_type"
447 @_ = ($WH{$type} or Carp::croak "unsupported type passed to AnyEvent::Handle::push_write") 1011 or Carp::croak "unsupported/unloadable type '$type' passed to AnyEvent::Handle::push_write")
448 ->($self, @_); 1012 ->($self, @_);
449 } 1013 }
450 1014
1015 # we downgrade here to avoid hard-to-track-down bugs,
1016 # and diagnose the problem earlier and better.
1017
451 if ($self->{filter_w}) { 1018 if ($self->{tls}) {
452 $self->{filter_w}($self, \$_[0]); 1019 utf8::downgrade $self->{_tls_wbuf} .= $_[0];
1020 &_dotls ($self) if $self->{fh};
453 } else { 1021 } else {
454 $self->{wbuf} .= $_[0]; 1022 utf8::downgrade $self->{wbuf} .= $_[0];
455 $self->_drain_wbuf; 1023 $self->_drain_wbuf if $self->{fh};
456 } 1024 }
457} 1025}
458 1026
459=item $handle->push_write (type => @args) 1027=item $handle->push_write (type => @args)
460 1028
461Instead of formatting your data yourself, you can also let this module do 1029Instead of formatting your data yourself, you can also let this module
462the job by specifying a type and type-specific arguments. 1030do the job by specifying a type and type-specific arguments. You
1031can also specify the (fully qualified) name of a package, in which
1032case AnyEvent tries to load the package and then expects to find the
1033C<anyevent_write_type> function inside (see "custom write types", below).
463 1034
464Predefined types are (if you have ideas for additional types, feel free to 1035Predefined types are (if you have ideas for additional types, feel free to
465drop by and tell us): 1036drop by and tell us):
466 1037
467=over 4 1038=over 4
474=cut 1045=cut
475 1046
476register_write_type netstring => sub { 1047register_write_type netstring => sub {
477 my ($self, $string) = @_; 1048 my ($self, $string) = @_;
478 1049
479 sprintf "%d:%s,", (length $string), $string 1050 (length $string) . ":$string,"
480}; 1051};
481 1052
482=item packstring => $format, $data 1053=item packstring => $format, $data
483 1054
484An octet string prefixed with an encoded length. The encoding C<$format> 1055An octet string prefixed with an encoded length. The encoding C<$format>
489=cut 1060=cut
490 1061
491register_write_type packstring => sub { 1062register_write_type packstring => sub {
492 my ($self, $format, $string) = @_; 1063 my ($self, $format, $string) = @_;
493 1064
494 pack "$format/a", $string 1065 pack "$format/a*", $string
495}; 1066};
496 1067
497=item json => $array_or_hashref 1068=item json => $array_or_hashref
498 1069
499Encodes the given hash or array reference into a JSON object. Unless you 1070Encodes the given hash or array reference into a JSON object. Unless you
500provide your own JSON object, this means it will be encoded to JSON text 1071provide your own JSON object, this means it will be encoded to JSON text
501in UTF-8. 1072in UTF-8.
502 1073
1074The default encoder might or might not handle every type of JSON value -
1075it might be limited to arrays and objects for security reasons. See the
1076C<json> constructor attribute for more details.
1077
503JSON objects (and arrays) are self-delimiting, so you can write JSON at 1078JSON objects (and arrays) are self-delimiting, so if you only use arrays
504one end of a handle and read them at the other end without using any 1079and hashes, you can write JSON at one end of a handle and read them at the
505additional framing. 1080other end without using any additional framing.
506 1081
507The generated JSON text is guaranteed not to contain any newlines: While 1082The JSON text generated by the default encoder is guaranteed not to
508this module doesn't need delimiters after or between JSON texts to be 1083contain any newlines: While this module doesn't need delimiters after or
509able to read them, many other languages depend on that. 1084between JSON texts to be able to read them, many other languages depend on
1085them.
510 1086
511A simple RPC protocol that interoperates easily with others is to send 1087A simple RPC protocol that interoperates easily with other languages is
512JSON arrays (or objects, although arrays are usually the better choice as 1088to send JSON arrays (or objects, although arrays are usually the better
513they mimic how function argument passing works) and a newline after each 1089choice as they mimic how function argument passing works) and a newline
514JSON text: 1090after each JSON text:
515 1091
516 $handle->push_write (json => ["method", "arg1", "arg2"]); # whatever 1092 $handle->push_write (json => ["method", "arg1", "arg2"]); # whatever
517 $handle->push_write ("\012"); 1093 $handle->push_write ("\012");
518 1094
519An AnyEvent::Handle receiver would simply use the C<json> read type and 1095An AnyEvent::Handle receiver would simply use the C<json> read type and
522 $handle->push_read (json => sub { my $array = $_[1]; ... }); 1098 $handle->push_read (json => sub { my $array = $_[1]; ... });
523 1099
524Other languages could read single lines terminated by a newline and pass 1100Other languages could read single lines terminated by a newline and pass
525this line into their JSON decoder of choice. 1101this line into their JSON decoder of choice.
526 1102
1103=item cbor => $perl_scalar
1104
1105Encodes the given scalar into a CBOR value. Unless you provide your own
1106L<CBOR::XS> object, this means it will be encoded to a CBOR string not
1107using any extensions, if possible.
1108
1109CBOR values are self-delimiting, so you can write CBOR at one end of
1110a handle and read them at the other end without using any additional
1111framing.
1112
1113A simple nd very very fast RPC protocol that interoperates with
1114other languages is to send CBOR and receive CBOR values (arrays are
1115recommended):
1116
1117 $handle->push_write (cbor => ["method", "arg1", "arg2"]); # whatever
1118
1119An AnyEvent::Handle receiver would simply use the C<cbor> read type:
1120
1121 $handle->push_read (cbor => sub { my $array = $_[1]; ... });
1122
527=cut 1123=cut
1124
1125sub json_coder() {
1126 eval { require JSON::XS; JSON::XS->new->utf8 }
1127 || do { require JSON::PP; JSON::PP->new->utf8 }
1128}
528 1129
529register_write_type json => sub { 1130register_write_type json => sub {
530 my ($self, $ref) = @_; 1131 my ($self, $ref) = @_;
531 1132
532 require JSON; 1133 ($self->{json} ||= json_coder)
533 1134 ->encode ($ref)
534 $self->{json} ? $self->{json}->encode ($ref)
535 : JSON::encode_json ($ref)
536}; 1135};
537 1136
1137sub cbor_coder() {
1138 require CBOR::XS;
1139 CBOR::XS->new
1140}
1141
1142register_write_type cbor => sub {
1143 my ($self, $scalar) = @_;
1144
1145 ($self->{cbor} ||= cbor_coder)
1146 ->encode ($scalar)
1147};
1148
1149=item storable => $reference
1150
1151Freezes the given reference using L<Storable> and writes it to the
1152handle. Uses the C<nfreeze> format.
1153
1154=cut
1155
1156register_write_type storable => sub {
1157 my ($self, $ref) = @_;
1158
1159 require Storable unless $Storable::VERSION;
1160
1161 pack "w/a*", Storable::nfreeze ($ref)
1162};
1163
538=back 1164=back
539 1165
540=item AnyEvent::Handle::register_write_type type => $coderef->($handle, @args) 1166=item $handle->push_shutdown
541 1167
542This function (not method) lets you add your own types to C<push_write>. 1168Sometimes you know you want to close the socket after writing your data
1169before it was actually written. One way to do that is to replace your
1170C<on_drain> handler by a callback that shuts down the socket (and set
1171C<low_water_mark> to C<0>). This method is a shorthand for just that, and
1172replaces the C<on_drain> callback with:
1173
1174 sub { shutdown $_[0]{fh}, 1 }
1175
1176This simply shuts down the write side and signals an EOF condition to the
1177the peer.
1178
1179You can rely on the normal read queue and C<on_eof> handling
1180afterwards. This is the cleanest way to close a connection.
1181
1182This method may invoke callbacks (and therefore the handle might be
1183destroyed after it returns).
1184
1185=cut
1186
1187sub push_shutdown {
1188 my ($self) = @_;
1189
1190 delete $self->{low_water_mark};
1191 $self->on_drain (sub { shutdown $_[0]{fh}, 1 });
1192}
1193
1194=item custom write types - Package::anyevent_write_type $handle, @args
1195
1196Instead of one of the predefined types, you can also specify the name of
1197a package. AnyEvent will try to load the package and then expects to find
1198a function named C<anyevent_write_type> inside. If it isn't found, it
1199progressively tries to load the parent package until it either finds the
1200function (good) or runs out of packages (bad).
1201
543Whenever the given C<type> is used, C<push_write> will invoke the code 1202Whenever the given C<type> is used, C<push_write> will the function with
544reference with the handle object and the remaining arguments. 1203the handle object and the remaining arguments.
545 1204
546The code reference is supposed to return a single octet string that will 1205The function is supposed to return a single octet string that will be
547be appended to the write buffer. 1206appended to the write buffer, so you can mentally treat this function as a
1207"arguments to on-the-wire-format" converter.
548 1208
549Note that this is a function, and all types registered this way will be 1209Example: implement a custom write type C<join> that joins the remaining
550global, so try to use unique names. 1210arguments using the first one.
1211
1212 $handle->push_write (My::Type => " ", 1,2,3);
1213
1214 # uses the following package, which can be defined in the "My::Type" or in
1215 # the "My" modules to be auto-loaded, or just about anywhere when the
1216 # My::Type::anyevent_write_type is defined before invoking it.
1217
1218 package My::Type;
1219
1220 sub anyevent_write_type {
1221 my ($handle, $delim, @args) = @_;
1222
1223 join $delim, @args
1224 }
551 1225
552=cut 1226=cut
553 1227
554############################################################################# 1228#############################################################################
555 1229
564ways, the "simple" way, using only C<on_read> and the "complex" way, using 1238ways, the "simple" way, using only C<on_read> and the "complex" way, using
565a queue. 1239a queue.
566 1240
567In the simple case, you just install an C<on_read> callback and whenever 1241In the simple case, you just install an C<on_read> callback and whenever
568new data arrives, it will be called. You can then remove some data (if 1242new data arrives, it will be called. You can then remove some data (if
569enough is there) from the read buffer (C<< $handle->rbuf >>) if you want 1243enough is there) from the read buffer (C<< $handle->rbuf >>). Or you can
570or not. 1244leave the data there if you want to accumulate more (e.g. when only a
1245partial message has been received so far), or change the read queue with
1246e.g. C<push_read>.
571 1247
572In the more complex case, you want to queue multiple callbacks. In this 1248In the more complex case, you want to queue multiple callbacks. In this
573case, AnyEvent::Handle will call the first queued callback each time new 1249case, AnyEvent::Handle will call the first queued callback each time new
574data arrives (also the first time it is queued) and removes it when it has 1250data arrives (also the first time it is queued) and remove it when it has
575done its job (see C<push_read>, below). 1251done its job (see C<push_read>, below).
576 1252
577This way you can, for example, push three line-reads, followed by reading 1253This way you can, for example, push three line-reads, followed by reading
578a chunk of data, and AnyEvent::Handle will execute them in order. 1254a chunk of data, and AnyEvent::Handle will execute them in order.
579 1255
593 # handle xml 1269 # handle xml
594 }); 1270 });
595 }); 1271 });
596 }); 1272 });
597 1273
598Example 2: Implement a client for a protocol that replies either with 1274Example 2: Implement a client for a protocol that replies either with "OK"
599"OK" and another line or "ERROR" for one request, and 64 bytes for the 1275and another line or "ERROR" for the first request that is sent, and 64
600second request. Due tot he availability of a full queue, we can just 1276bytes for the second request. Due to the availability of a queue, we can
601pipeline sending both requests and manipulate the queue as necessary in 1277just pipeline sending both requests and manipulate the queue as necessary
602the callbacks: 1278in the callbacks.
603 1279
604 # request one 1280When the first callback is called and sees an "OK" response, it will
1281C<unshift> another line-read. This line-read will be queued I<before> the
128264-byte chunk callback.
1283
1284 # request one, returns either "OK + extra line" or "ERROR"
605 $handle->push_write ("request 1\015\012"); 1285 $handle->push_write ("request 1\015\012");
606 1286
607 # we expect "ERROR" or "OK" as response, so push a line read 1287 # we expect "ERROR" or "OK" as response, so push a line read
608 $handle->push_read (line => sub { 1288 $handle->push_read (line => sub {
609 # if we got an "OK", we have to _prepend_ another line, 1289 # if we got an "OK", we have to _prepend_ another line,
616 ... 1296 ...
617 }); 1297 });
618 } 1298 }
619 }); 1299 });
620 1300
621 # request two 1301 # request two, simply returns 64 octets
622 $handle->push_write ("request 2\015\012"); 1302 $handle->push_write ("request 2\015\012");
623 1303
624 # simply read 64 bytes, always 1304 # simply read 64 bytes, always
625 $handle->push_read (chunk => 64, sub { 1305 $handle->push_read (chunk => 64, sub {
626 my $response = $_[1]; 1306 my $response = $_[1];
632=cut 1312=cut
633 1313
634sub _drain_rbuf { 1314sub _drain_rbuf {
635 my ($self) = @_; 1315 my ($self) = @_;
636 1316
1317 # avoid recursion
1318 return if $self->{_skip_drain_rbuf};
637 local $self->{_in_drain} = 1; 1319 local $self->{_skip_drain_rbuf} = 1;
638
639 if (
640 defined $self->{rbuf_max}
641 && $self->{rbuf_max} < length $self->{rbuf}
642 ) {
643 return $self->_error (&Errno::ENOSPC, 1);
644 }
645 1320
646 while () { 1321 while () {
647 no strict 'refs'; 1322 # we need to use a separate tls read buffer, as we must not receive data while
1323 # we are draining the buffer, and this can only happen with TLS.
1324 $self->{rbuf} .= delete $self->{_tls_rbuf}
1325 if exists $self->{_tls_rbuf};
648 1326
649 my $len = length $self->{rbuf}; 1327 my $len = length $self->{rbuf};
650 1328
651 if (my $cb = shift @{ $self->{_queue} }) { 1329 if (my $cb = shift @{ $self->{_queue} }) {
652 unless ($cb->($self)) { 1330 unless ($cb->($self)) {
653 if ($self->{_eof}) { 1331 # no progress can be made
654 # no progress can be made (not enough data and no data forthcoming) 1332 # (not enough data and no data forthcoming)
655 $self->_error (&Errno::EPIPE, 1), last; 1333 $self->_error (Errno::EPIPE, 1), return
656 } 1334 if $self->{_eof};
657 1335
658 unshift @{ $self->{_queue} }, $cb; 1336 unshift @{ $self->{_queue} }, $cb;
659 last; 1337 last;
660 } 1338 }
661 } elsif ($self->{on_read}) { 1339 } elsif ($self->{on_read}) {
668 && !@{ $self->{_queue} } # and the queue is still empty 1346 && !@{ $self->{_queue} } # and the queue is still empty
669 && $self->{on_read} # but we still have on_read 1347 && $self->{on_read} # but we still have on_read
670 ) { 1348 ) {
671 # no further data will arrive 1349 # no further data will arrive
672 # so no progress can be made 1350 # so no progress can be made
673 $self->_error (&Errno::EPIPE, 1), last 1351 $self->_error (Errno::EPIPE, 1), return
674 if $self->{_eof}; 1352 if $self->{_eof};
675 1353
676 last; # more data might arrive 1354 last; # more data might arrive
677 } 1355 }
678 } else { 1356 } else {
679 # read side becomes idle 1357 # read side becomes idle
680 delete $self->{_rw}; 1358 delete $self->{_rw} unless $self->{tls};
681 last; 1359 last;
682 } 1360 }
683 } 1361 }
684 1362
1363 if ($self->{_eof}) {
1364 $self->{on_eof}
685 $self->{on_eof}($self) 1365 ? $self->{on_eof}($self)
686 if $self->{_eof} && $self->{on_eof}; 1366 : $self->_error (0, 1, "Unexpected end-of-file");
1367
1368 return;
1369 }
1370
1371 if (
1372 defined $self->{rbuf_max}
1373 && $self->{rbuf_max} < length $self->{rbuf}
1374 ) {
1375 $self->_error (Errno::ENOSPC, 1), return;
1376 }
687 1377
688 # may need to restart read watcher 1378 # may need to restart read watcher
689 unless ($self->{_rw}) { 1379 unless ($self->{_rw}) {
690 $self->start_read 1380 $self->start_read
691 if $self->{on_read} || @{ $self->{_queue} }; 1381 if $self->{on_read} || @{ $self->{_queue} };
696 1386
697This replaces the currently set C<on_read> callback, or clears it (when 1387This replaces the currently set C<on_read> callback, or clears it (when
698the new callback is C<undef>). See the description of C<on_read> in the 1388the new callback is C<undef>). See the description of C<on_read> in the
699constructor. 1389constructor.
700 1390
1391This method may invoke callbacks (and therefore the handle might be
1392destroyed after it returns).
1393
701=cut 1394=cut
702 1395
703sub on_read { 1396sub on_read {
704 my ($self, $cb) = @_; 1397 my ($self, $cb) = @_;
705 1398
706 $self->{on_read} = $cb; 1399 $self->{on_read} = $cb;
707 $self->_drain_rbuf if $cb && !$self->{_in_drain}; 1400 $self->_drain_rbuf if $cb;
708} 1401}
709 1402
710=item $handle->rbuf 1403=item $handle->rbuf
711 1404
712Returns the read buffer (as a modifiable lvalue). 1405Returns the read buffer (as a modifiable lvalue). You can also access the
1406read buffer directly as the C<< ->{rbuf} >> member, if you want (this is
1407much faster, and no less clean).
713 1408
714You can access the read buffer directly as the C<< ->{rbuf} >> member, if 1409The only operation allowed on the read buffer (apart from looking at it)
715you want. 1410is removing data from its beginning. Otherwise modifying or appending to
1411it is not allowed and will lead to hard-to-track-down bugs.
716 1412
717NOTE: The read buffer should only be used or modified if the C<on_read>, 1413NOTE: The read buffer should only be used or modified in the C<on_read>
718C<push_read> or C<unshift_read> methods are used. The other read methods 1414callback or when C<push_read> or C<unshift_read> are used with a single
719automatically manage the read buffer. 1415callback (i.e. untyped). Typed C<push_read> and C<unshift_read> methods
1416will manage the read buffer on their own.
720 1417
721=cut 1418=cut
722 1419
723sub rbuf : lvalue { 1420sub rbuf : lvalue {
724 $_[0]{rbuf} 1421 $_[0]{rbuf}
741 1438
742If enough data was available, then the callback must remove all data it is 1439If enough data was available, then the callback must remove all data it is
743interested in (which can be none at all) and return a true value. After returning 1440interested in (which can be none at all) and return a true value. After returning
744true, it will be removed from the queue. 1441true, it will be removed from the queue.
745 1442
1443These methods may invoke callbacks (and therefore the handle might be
1444destroyed after it returns).
1445
746=cut 1446=cut
747 1447
748our %RH; 1448our %RH;
749 1449
750sub register_read_type($$) { 1450sub register_read_type($$) {
756 my $cb = pop; 1456 my $cb = pop;
757 1457
758 if (@_) { 1458 if (@_) {
759 my $type = shift; 1459 my $type = shift;
760 1460
1461 $cb = ($RH{$type} ||= _load_func "$type\::anyevent_read_type"
761 $cb = ($RH{$type} or Carp::croak "unsupported type passed to AnyEvent::Handle::push_read") 1462 or Carp::croak "unsupported/unloadable type '$type' passed to AnyEvent::Handle::push_read")
762 ->($self, $cb, @_); 1463 ->($self, $cb, @_);
763 } 1464 }
764 1465
765 push @{ $self->{_queue} }, $cb; 1466 push @{ $self->{_queue} }, $cb;
766 $self->_drain_rbuf unless $self->{_in_drain}; 1467 $self->_drain_rbuf;
767} 1468}
768 1469
769sub unshift_read { 1470sub unshift_read {
770 my $self = shift; 1471 my $self = shift;
771 my $cb = pop; 1472 my $cb = pop;
772 1473
773 if (@_) { 1474 if (@_) {
774 my $type = shift; 1475 my $type = shift;
775 1476
1477 $cb = ($RH{$type} ||= _load_func "$type\::anyevent_read_type"
776 $cb = ($RH{$type} or Carp::croak "unsupported type passed to AnyEvent::Handle::unshift_read") 1478 or Carp::croak "unsupported/unloadable type '$type' passed to AnyEvent::Handle::unshift_read")
777 ->($self, $cb, @_); 1479 ->($self, $cb, @_);
778 } 1480 }
779 1481
780
781 unshift @{ $self->{_queue} }, $cb; 1482 unshift @{ $self->{_queue} }, $cb;
782 $self->_drain_rbuf unless $self->{_in_drain}; 1483 $self->_drain_rbuf;
783} 1484}
784 1485
785=item $handle->push_read (type => @args, $cb) 1486=item $handle->push_read (type => @args, $cb)
786 1487
787=item $handle->unshift_read (type => @args, $cb) 1488=item $handle->unshift_read (type => @args, $cb)
788 1489
789Instead of providing a callback that parses the data itself you can chose 1490Instead of providing a callback that parses the data itself you can chose
790between a number of predefined parsing formats, for chunks of data, lines 1491between a number of predefined parsing formats, for chunks of data, lines
791etc. 1492etc. You can also specify the (fully qualified) name of a package, in
1493which case AnyEvent tries to load the package and then expects to find the
1494C<anyevent_read_type> function inside (see "custom read types", below).
792 1495
793Predefined types are (if you have ideas for additional types, feel free to 1496Predefined types are (if you have ideas for additional types, feel free to
794drop by and tell us): 1497drop by and tell us):
795 1498
796=over 4 1499=over 4
802data. 1505data.
803 1506
804Example: read 2 bytes. 1507Example: read 2 bytes.
805 1508
806 $handle->push_read (chunk => 2, sub { 1509 $handle->push_read (chunk => 2, sub {
807 warn "yay ", unpack "H*", $_[1]; 1510 say "yay " . unpack "H*", $_[1];
808 }); 1511 });
809 1512
810=cut 1513=cut
811 1514
812register_read_type chunk => sub { 1515register_read_type chunk => sub {
817 $cb->($_[0], substr $_[0]{rbuf}, 0, $len, ""); 1520 $cb->($_[0], substr $_[0]{rbuf}, 0, $len, "");
818 1 1521 1
819 } 1522 }
820}; 1523};
821 1524
822# compatibility with older API
823sub push_read_chunk {
824 $_[0]->push_read (chunk => $_[1], $_[2]);
825}
826
827sub unshift_read_chunk {
828 $_[0]->unshift_read (chunk => $_[1], $_[2]);
829}
830
831=item line => [$eol, ]$cb->($handle, $line, $eol) 1525=item line => [$eol, ]$cb->($handle, $line, $eol)
832 1526
833The callback will be called only once a full line (including the end of 1527The callback will be called only once a full line (including the end of
834line marker, C<$eol>) has been read. This line (excluding the end of line 1528line marker, C<$eol>) has been read. This line (excluding the end of line
835marker) will be passed to the callback as second argument (C<$line>), and 1529marker) will be passed to the callback as second argument (C<$line>), and
850=cut 1544=cut
851 1545
852register_read_type line => sub { 1546register_read_type line => sub {
853 my ($self, $cb, $eol) = @_; 1547 my ($self, $cb, $eol) = @_;
854 1548
855 $eol = qr|(\015?\012)| if @_ < 3; 1549 if (@_ < 3) {
1550 # this is faster then the generic code below
1551 sub {
1552 (my $pos = index $_[0]{rbuf}, "\012") >= 0
1553 or return;
1554
1555 (my $str = substr $_[0]{rbuf}, 0, $pos + 1, "") =~ s/(\015?\012)\Z// or die;
1556 $cb->($_[0], $str, "$1");
1557 1
1558 }
1559 } else {
856 $eol = quotemeta $eol unless ref $eol; 1560 $eol = quotemeta $eol unless ref $eol;
857 $eol = qr|^(.*?)($eol)|s; 1561 $eol = qr|^(.*?)($eol)|s;
858 1562
859 sub { 1563 sub {
860 $_[0]{rbuf} =~ s/$eol// or return; 1564 $_[0]{rbuf} =~ s/$eol// or return;
861 1565
862 $cb->($_[0], $1, $2); 1566 $cb->($_[0], "$1", "$2");
1567 1
863 1 1568 }
864 } 1569 }
865}; 1570};
866
867# compatibility with older API
868sub push_read_line {
869 my $self = shift;
870 $self->push_read (line => @_);
871}
872
873sub unshift_read_line {
874 my $self = shift;
875 $self->unshift_read (line => @_);
876}
877 1571
878=item regex => $accept[, $reject[, $skip], $cb->($handle, $data) 1572=item regex => $accept[, $reject[, $skip], $cb->($handle, $data)
879 1573
880Makes a regex match against the regex object C<$accept> and returns 1574Makes a regex match against the regex object C<$accept> and returns
881everything up to and including the match. 1575everything up to and including the match.
899the receive buffer when neither C<$accept> nor C<$reject> match, 1593the receive buffer when neither C<$accept> nor C<$reject> match,
900and everything preceding and including the match will be accepted 1594and everything preceding and including the match will be accepted
901unconditionally. This is useful to skip large amounts of data that you 1595unconditionally. This is useful to skip large amounts of data that you
902know cannot be matched, so that the C<$accept> or C<$reject> regex do not 1596know cannot be matched, so that the C<$accept> or C<$reject> regex do not
903have to start matching from the beginning. This is purely an optimisation 1597have to start matching from the beginning. This is purely an optimisation
904and is usually worth only when you expect more than a few kilobytes. 1598and is usually worth it only when you expect more than a few kilobytes.
905 1599
906Example: expect a http header, which ends at C<\015\012\015\012>. Since we 1600Example: expect a http header, which ends at C<\015\012\015\012>. Since we
907expect the header to be very large (it isn't in practise, but...), we use 1601expect the header to be very large (it isn't in practice, but...), we use
908a skip regex to skip initial portions. The skip regex is tricky in that 1602a skip regex to skip initial portions. The skip regex is tricky in that
909it only accepts something not ending in either \015 or \012, as these are 1603it only accepts something not ending in either \015 or \012, as these are
910required for the accept regex. 1604required for the accept regex.
911 1605
912 $handle->push_read (regex => 1606 $handle->push_read (regex =>
925 1619
926 sub { 1620 sub {
927 # accept 1621 # accept
928 if ($$rbuf =~ $accept) { 1622 if ($$rbuf =~ $accept) {
929 $data .= substr $$rbuf, 0, $+[0], ""; 1623 $data .= substr $$rbuf, 0, $+[0], "";
930 $cb->($self, $data); 1624 $cb->($_[0], $data);
931 return 1; 1625 return 1;
932 } 1626 }
933 1627
934 # reject 1628 # reject
935 if ($reject && $$rbuf =~ $reject) { 1629 if ($reject && $$rbuf =~ $reject) {
936 $self->_error (&Errno::EBADMSG); 1630 $_[0]->_error (Errno::EBADMSG);
937 } 1631 }
938 1632
939 # skip 1633 # skip
940 if ($skip && $$rbuf =~ $skip) { 1634 if ($skip && $$rbuf =~ $skip) {
941 $data .= substr $$rbuf, 0, $+[0], ""; 1635 $data .= substr $$rbuf, 0, $+[0], "";
957 my ($self, $cb) = @_; 1651 my ($self, $cb) = @_;
958 1652
959 sub { 1653 sub {
960 unless ($_[0]{rbuf} =~ s/^(0|[1-9][0-9]*)://) { 1654 unless ($_[0]{rbuf} =~ s/^(0|[1-9][0-9]*)://) {
961 if ($_[0]{rbuf} =~ /[^0-9]/) { 1655 if ($_[0]{rbuf} =~ /[^0-9]/) {
962 $self->_error (&Errno::EBADMSG); 1656 $_[0]->_error (Errno::EBADMSG);
963 } 1657 }
964 return; 1658 return;
965 } 1659 }
966 1660
967 my $len = $1; 1661 my $len = $1;
968 1662
969 $self->unshift_read (chunk => $len, sub { 1663 $_[0]->unshift_read (chunk => $len, sub {
970 my $string = $_[1]; 1664 my $string = $_[1];
971 $_[0]->unshift_read (chunk => 1, sub { 1665 $_[0]->unshift_read (chunk => 1, sub {
972 if ($_[1] eq ",") { 1666 if ($_[1] eq ",") {
973 $cb->($_[0], $string); 1667 $cb->($_[0], $string);
974 } else { 1668 } else {
975 $self->_error (&Errno::EBADMSG); 1669 $_[0]->_error (Errno::EBADMSG);
976 } 1670 }
977 }); 1671 });
978 }); 1672 });
979 1673
980 1 1674 1
986An octet string prefixed with an encoded length. The encoding C<$format> 1680An octet string prefixed with an encoded length. The encoding C<$format>
987uses the same format as a Perl C<pack> format, but must specify a single 1681uses the same format as a Perl C<pack> format, but must specify a single
988integer only (only one of C<cCsSlLqQiInNvVjJw> is allowed, plus an 1682integer only (only one of C<cCsSlLqQiInNvVjJw> is allowed, plus an
989optional C<!>, C<< < >> or C<< > >> modifier). 1683optional C<!>, C<< < >> or C<< > >> modifier).
990 1684
991DNS over TCP uses a prefix of C<n>, EPP uses a prefix of C<N>. 1685For example, DNS over TCP uses a prefix of C<n> (2 octet network order),
1686EPP uses a prefix of C<N> (4 octtes).
992 1687
993Example: read a block of data prefixed by its length in BER-encoded 1688Example: read a block of data prefixed by its length in BER-encoded
994format (very efficient). 1689format (very efficient).
995 1690
996 $handle->push_read (packstring => "w", sub { 1691 $handle->push_read (packstring => "w", sub {
1002register_read_type packstring => sub { 1697register_read_type packstring => sub {
1003 my ($self, $cb, $format) = @_; 1698 my ($self, $cb, $format) = @_;
1004 1699
1005 sub { 1700 sub {
1006 # when we can use 5.10 we can use ".", but for 5.8 we use the re-pack method 1701 # when we can use 5.10 we can use ".", but for 5.8 we use the re-pack method
1007 defined (my $len = eval { unpack $format, $_[0]->{rbuf} }) 1702 defined (my $len = eval { unpack $format, $_[0]{rbuf} })
1008 or return; 1703 or return;
1009 1704
1705 $format = length pack $format, $len;
1706
1707 # bypass unshift if we already have the remaining chunk
1708 if ($format + $len <= length $_[0]{rbuf}) {
1709 my $data = substr $_[0]{rbuf}, $format, $len;
1710 substr $_[0]{rbuf}, 0, $format + $len, "";
1711 $cb->($_[0], $data);
1712 } else {
1010 # remove prefix 1713 # remove prefix
1011 substr $_[0]->{rbuf}, 0, (length pack $format, $len), ""; 1714 substr $_[0]{rbuf}, 0, $format, "";
1012 1715
1013 # read rest 1716 # read remaining chunk
1014 $_[0]->unshift_read (chunk => $len, $cb); 1717 $_[0]->unshift_read (chunk => $len, $cb);
1718 }
1015 1719
1016 1 1720 1
1017 } 1721 }
1018}; 1722};
1019 1723
1020=item json => $cb->($handle, $hash_or_arrayref) 1724=item json => $cb->($handle, $hash_or_arrayref)
1021 1725
1022Reads a JSON object or array, decodes it and passes it to the callback. 1726Reads a JSON object or array, decodes it and passes it to the
1727callback. When a parse error occurs, an C<EBADMSG> error will be raised.
1023 1728
1024If a C<json> object was passed to the constructor, then that will be used 1729If a C<json> object was passed to the constructor, then that will be
1025for the final decode, otherwise it will create a JSON coder expecting UTF-8. 1730used for the final decode, otherwise it will create a L<JSON::XS> or
1731L<JSON::PP> coder object expecting UTF-8.
1026 1732
1027This read type uses the incremental parser available with JSON version 1733This read type uses the incremental parser available with JSON version
10282.09 (and JSON::XS version 2.2) and above. You have to provide a 17342.09 (and JSON::XS version 2.2) and above.
1029dependency on your own: this module will load the JSON module, but
1030AnyEvent does not depend on it itself.
1031 1735
1032Since JSON texts are fully self-delimiting, the C<json> read and write 1736Since JSON texts are fully self-delimiting, the C<json> read and write
1033types are an ideal simple RPC protocol: just exchange JSON datagrams. See 1737types are an ideal simple RPC protocol: just exchange JSON datagrams. See
1034the C<json> write type description, above, for an actual example. 1738the C<json> write type description, above, for an actual example.
1035 1739
1036=cut 1740=cut
1037 1741
1038register_read_type json => sub { 1742register_read_type json => sub {
1039 my ($self, $cb, $accept, $reject, $skip) = @_; 1743 my ($self, $cb) = @_;
1040 1744
1041 require JSON; 1745 my $json = $self->{json} ||= json_coder;
1042 1746
1043 my $data; 1747 my $data;
1044 my $rbuf = \$self->{rbuf};
1045
1046 my $json = $self->{json} ||= JSON->new->utf8;
1047 1748
1048 sub { 1749 sub {
1049 my $ref = $json->incr_parse ($self->{rbuf}); 1750 my $ref = eval { $json->incr_parse ($_[0]{rbuf}) };
1050 1751
1051 if ($ref) { 1752 if ($ref) {
1052 $self->{rbuf} = $json->incr_text; 1753 $_[0]{rbuf} = $json->incr_text;
1053 $json->incr_text = ""; 1754 $json->incr_text = "";
1054 $cb->($self, $ref); 1755 $cb->($_[0], $ref);
1055 1756
1056 1 1757 1
1758 } elsif ($@) {
1759 # error case
1760 $json->incr_skip;
1761
1762 $_[0]{rbuf} = $json->incr_text;
1763 $json->incr_text = "";
1764
1765 $_[0]->_error (Errno::EBADMSG);
1766
1767 ()
1057 } else { 1768 } else {
1058 $self->{rbuf} = ""; 1769 $_[0]{rbuf} = "";
1770
1059 () 1771 ()
1060 } 1772 }
1061 } 1773 }
1062}; 1774};
1063 1775
1776=item cbor => $cb->($handle, $scalar)
1777
1778Reads a CBOR value, decodes it and passes it to the callback. When a parse
1779error occurs, an C<EBADMSG> error will be raised.
1780
1781If a L<CBOR::XS> object was passed to the constructor, then that will be
1782used for the final decode, otherwise it will create a CBOR coder without
1783enabling any options.
1784
1785You have to provide a dependency to L<CBOR::XS> on your own: this module
1786will load the L<CBOR::XS> module, but AnyEvent does not depend on it
1787itself.
1788
1789Since CBOR values are fully self-delimiting, the C<cbor> read and write
1790types are an ideal simple RPC protocol: just exchange CBOR datagrams. See
1791the C<cbor> write type description, above, for an actual example.
1792
1793=cut
1794
1795register_read_type cbor => sub {
1796 my ($self, $cb) = @_;
1797
1798 my $cbor = $self->{cbor} ||= cbor_coder;
1799
1800 my $data;
1801
1802 sub {
1803 my (@value) = eval { $cbor->incr_parse ($_[0]{rbuf}) };
1804
1805 if (@value) {
1806 $cb->($_[0], @value);
1807
1808 1
1809 } elsif ($@) {
1810 # error case
1811 $cbor->incr_reset;
1812
1813 $_[0]->_error (Errno::EBADMSG);
1814
1815 ()
1816 } else {
1817 ()
1818 }
1819 }
1820};
1821
1822=item storable => $cb->($handle, $ref)
1823
1824Deserialises a L<Storable> frozen representation as written by the
1825C<storable> write type (BER-encoded length prefix followed by nfreeze'd
1826data).
1827
1828Raises C<EBADMSG> error if the data could not be decoded.
1829
1830=cut
1831
1832register_read_type storable => sub {
1833 my ($self, $cb) = @_;
1834
1835 require Storable unless $Storable::VERSION;
1836
1837 sub {
1838 # when we can use 5.10 we can use ".", but for 5.8 we use the re-pack method
1839 defined (my $len = eval { unpack "w", $_[0]{rbuf} })
1840 or return;
1841
1842 my $format = length pack "w", $len;
1843
1844 # bypass unshift if we already have the remaining chunk
1845 if ($format + $len <= length $_[0]{rbuf}) {
1846 my $data = substr $_[0]{rbuf}, $format, $len;
1847 substr $_[0]{rbuf}, 0, $format + $len, "";
1848
1849 eval { $cb->($_[0], Storable::thaw ($data)); 1 }
1850 or return $_[0]->_error (Errno::EBADMSG);
1851 } else {
1852 # remove prefix
1853 substr $_[0]{rbuf}, 0, $format, "";
1854
1855 # read remaining chunk
1856 $_[0]->unshift_read (chunk => $len, sub {
1857 eval { $cb->($_[0], Storable::thaw ($_[1])); 1 }
1858 or $_[0]->_error (Errno::EBADMSG);
1859 });
1860 }
1861
1862 1
1863 }
1864};
1865
1866=item tls_detect => $cb->($handle, $detect, $major, $minor)
1867
1868Checks the input stream for a valid SSL or TLS handshake TLSPaintext
1869record without consuming anything. Only SSL version 3 or higher
1870is handled, up to the fictituous protocol 4.x (but both SSL3+ and
1871SSL2-compatible framing is supported).
1872
1873If it detects that the input data is likely TLS, it calls the callback
1874with a true value for C<$detect> and the (on-wire) TLS version as second
1875and third argument (C<$major> is C<3>, and C<$minor> is 0..3 for SSL
18763.0, TLS 1.0, 1.1 and 1.2, respectively). If it detects the input to
1877be definitely not TLS, it calls the callback with a false value for
1878C<$detect>.
1879
1880The callback could use this information to decide whether or not to start
1881TLS negotiation.
1882
1883In all cases the data read so far is passed to the following read
1884handlers.
1885
1886Usually you want to use the C<tls_autostart> read type instead.
1887
1888If you want to design a protocol that works in the presence of TLS
1889dtection, make sure that any non-TLS data doesn't start with the octet 22
1890(ASCII SYN, 16 hex) or 128-255 (i.e. highest bit set). The checks this
1891read type does are a bit more strict, but might losen in the future to
1892accomodate protocol changes.
1893
1894This read type does not rely on L<AnyEvent::TLS> (and thus, not on
1895L<Net::SSLeay>).
1896
1897=item tls_autostart => $tls[, $tls_ctx]
1898
1899Tries to detect a valid SSL or TLS handshake. If one is detected, it tries
1900to start tls by calling C<starttls> with the given arguments.
1901
1902In practise, C<$tls> must be C<accept>, or a Net::SSLeay context that has
1903been configured to accept, as servers do not normally send a handshake on
1904their own and ths cannot be detected in this way.
1905
1906See C<tls_detect> above for more details.
1907
1908Example: give the client a chance to start TLS before accepting a text
1909line.
1910
1911 $hdl->push_read (tls_detect => "accept");
1912 $hdl->push_read (line => sub {
1913 print "received ", ($_[0]{tls} ? "encrypted" : "cleartext"), " <$_[1]>\n";
1914 });
1915
1916=cut
1917
1918register_read_type tls_detect => sub {
1919 my ($self, $cb) = @_;
1920
1921 sub {
1922 # this regex matches a full or partial tls record
1923 if (
1924 # ssl3+: type(22=handshake) major(=3) minor(any) length_hi
1925 $self->{rbuf} =~ /^(?:\z| \x16 (\z| [\x03\x04] (?:\z| . (?:\z| [\x00-\x40] ))))/xs
1926 # ssl2 comapatible: len_hi len_lo type(1) major minor dummy(forlength)
1927 or $self->{rbuf} =~ /^(?:\z| [\x80-\xff] (?:\z| . (?:\z| \x01 (\z| [\x03\x04] (?:\z| . (?:\z| . ))))))/xs
1928 ) {
1929 return if 3 != length $1; # partial match, can't decide yet
1930
1931 # full match, valid TLS record
1932 my ($major, $minor) = unpack "CC", $1;
1933 $cb->($self, "accept", $major + $minor * 0.1);
1934 } else {
1935 # mismatch == guaranteed not TLS
1936 $cb->($self, undef);
1937 }
1938
1939 1
1940 }
1941};
1942
1943register_read_type tls_autostart => sub {
1944 my ($self, @tls) = @_;
1945
1946 $RH{tls_detect}($self, sub {
1947 return unless $_[1];
1948 $_[0]->starttls (@tls);
1949 })
1950};
1951
1064=back 1952=back
1065 1953
1066=item AnyEvent::Handle::register_read_type type => $coderef->($handle, $cb, @args) 1954=item custom read types - Package::anyevent_read_type $handle, $cb, @args
1067 1955
1068This function (not method) lets you add your own types to C<push_read>. 1956Instead of one of the predefined types, you can also specify the name
1957of a package. AnyEvent will try to load the package and then expects to
1958find a function named C<anyevent_read_type> inside. If it isn't found, it
1959progressively tries to load the parent package until it either finds the
1960function (good) or runs out of packages (bad).
1069 1961
1070Whenever the given C<type> is used, C<push_read> will invoke the code 1962Whenever this type is used, C<push_read> will invoke the function with the
1071reference with the handle object, the callback and the remaining 1963handle object, the original callback and the remaining arguments.
1072arguments.
1073 1964
1074The code reference is supposed to return a callback (usually a closure) 1965The function is supposed to return a callback (usually a closure) that
1075that works as a plain read callback (see C<< ->push_read ($cb) >>). 1966works as a plain read callback (see C<< ->push_read ($cb) >>), so you can
1967mentally treat the function as a "configurable read type to read callback"
1968converter.
1076 1969
1077It should invoke the passed callback when it is done reading (remember to 1970It should invoke the original callback when it is done reading (remember
1078pass C<$handle> as first argument as all other callbacks do that). 1971to pass C<$handle> as first argument as all other callbacks do that,
1972although there is no strict requirement on this).
1079 1973
1080Note that this is a function, and all types registered this way will be
1081global, so try to use unique names.
1082
1083For examples, see the source of this module (F<perldoc -m AnyEvent::Handle>, 1974For examples, see the source of this module (F<perldoc -m
1084search for C<register_read_type>)). 1975AnyEvent::Handle>, search for C<register_read_type>)).
1085 1976
1086=item $handle->stop_read 1977=item $handle->stop_read
1087 1978
1088=item $handle->start_read 1979=item $handle->start_read
1089 1980
1095Note that AnyEvent::Handle will automatically C<start_read> for you when 1986Note that AnyEvent::Handle will automatically C<start_read> for you when
1096you change the C<on_read> callback or push/unshift a read callback, and it 1987you change the C<on_read> callback or push/unshift a read callback, and it
1097will automatically C<stop_read> for you when neither C<on_read> is set nor 1988will automatically C<stop_read> for you when neither C<on_read> is set nor
1098there are any read requests in the queue. 1989there are any read requests in the queue.
1099 1990
1991In older versions of this module (<= 5.3), these methods had no effect,
1992as TLS does not support half-duplex connections. In current versions they
1993work as expected, as this behaviour is required to avoid certain resource
1994attacks, where the program would be forced to read (and buffer) arbitrary
1995amounts of data before being able to send some data. The drawback is that
1996some readings of the the SSL/TLS specifications basically require this
1997attack to be working, as SSL/TLS implementations might stall sending data
1998during a rehandshake.
1999
2000As a guideline, during the initial handshake, you should not stop reading,
2001and as a client, it might cause problems, depending on your application.
2002
1100=cut 2003=cut
1101 2004
1102sub stop_read { 2005sub stop_read {
1103 my ($self) = @_; 2006 my ($self) = @_;
1104 2007
1106} 2009}
1107 2010
1108sub start_read { 2011sub start_read {
1109 my ($self) = @_; 2012 my ($self) = @_;
1110 2013
1111 unless ($self->{_rw} || $self->{_eof}) { 2014 unless ($self->{_rw} || $self->{_eof} || !$self->{fh}) {
1112 Scalar::Util::weaken $self; 2015 Scalar::Util::weaken $self;
1113 2016
1114 $self->{_rw} = AnyEvent->io (fh => $self->{fh}, poll => "r", cb => sub { 2017 $self->{_rw} = AE::io $self->{fh}, 0, sub {
1115 my $rbuf = $self->{filter_r} ? \my $buf : \$self->{rbuf}; 2018 my $rbuf = \($self->{tls} ? my $buf : $self->{rbuf});
1116 my $len = sysread $self->{fh}, $$rbuf, $self->{read_size} || 8192, length $$rbuf; 2019 my $len = sysread $self->{fh}, $$rbuf, $self->{read_size}, length $$rbuf;
1117 2020
1118 if ($len > 0) { 2021 if ($len > 0) {
1119 $self->{_activity} = AnyEvent->now; 2022 $self->{_activity} = $self->{_ractivity} = AE::now;
1120 2023
1121 $self->{filter_r} 2024 if ($self->{tls}) {
1122 ? $self->{filter_r}($self, $rbuf) 2025 Net::SSLeay::BIO_write ($self->{_rbio}, $$rbuf);
1123 : $self->{_in_drain} || $self->_drain_rbuf; 2026
2027 &_dotls ($self);
2028 } else {
2029 $self->_drain_rbuf;
2030 }
2031
2032 if ($len == $self->{read_size}) {
2033 $self->{read_size} *= 2;
2034 $self->{read_size} = $self->{max_read_size} || MAX_READ_SIZE
2035 if $self->{read_size} > ($self->{max_read_size} || MAX_READ_SIZE);
2036 }
1124 2037
1125 } elsif (defined $len) { 2038 } elsif (defined $len) {
1126 delete $self->{_rw}; 2039 delete $self->{_rw};
1127 $self->{_eof} = 1; 2040 $self->{_eof} = 1;
1128 $self->_drain_rbuf unless $self->{_in_drain}; 2041 $self->_drain_rbuf;
1129 2042
1130 } elsif ($! != EAGAIN && $! != EINTR && $! != WSAEWOULDBLOCK) { 2043 } elsif ($! != EAGAIN && $! != EINTR && $! != EWOULDBLOCK && $! != WSAEWOULDBLOCK) {
1131 return $self->_error ($!, 1); 2044 return $self->_error ($!, 1);
1132 } 2045 }
1133 }); 2046 };
1134 } 2047 }
1135} 2048}
1136 2049
2050our $ERROR_SYSCALL;
2051our $ERROR_WANT_READ;
2052
2053sub _tls_error {
2054 my ($self, $err) = @_;
2055
2056 return $self->_error ($!, 1)
2057 if $err == Net::SSLeay::ERROR_SYSCALL ();
2058
2059 my $err = Net::SSLeay::ERR_error_string (Net::SSLeay::ERR_get_error ());
2060
2061 # reduce error string to look less scary
2062 $err =~ s/^error:[0-9a-fA-F]{8}:[^:]+:([^:]+):/\L$1: /;
2063
2064 if ($self->{_on_starttls}) {
2065 (delete $self->{_on_starttls})->($self, undef, $err);
2066 &_freetls;
2067 } else {
2068 &_freetls;
2069 $self->_error (Errno::EPROTO, 1, $err);
2070 }
2071}
2072
2073# poll the write BIO and send the data if applicable
2074# also decode read data if possible
2075# this is basiclaly our TLS state machine
2076# more efficient implementations are possible with openssl,
2077# but not with the buggy and incomplete Net::SSLeay.
1137sub _dotls { 2078sub _dotls {
1138 my ($self) = @_; 2079 my ($self) = @_;
1139 2080
1140 my $buf; 2081 my $tmp;
1141 2082
1142 if (length $self->{_tls_wbuf}) { 2083 while (length $self->{_tls_wbuf}) {
1143 while ((my $len = Net::SSLeay::write ($self->{tls}, $self->{_tls_wbuf})) > 0) { 2084 if (($tmp = Net::SSLeay::write ($self->{tls}, $self->{_tls_wbuf})) <= 0) {
1144 substr $self->{_tls_wbuf}, 0, $len, ""; 2085 $tmp = Net::SSLeay::get_error ($self->{tls}, $tmp);
2086
2087 return $self->_tls_error ($tmp)
2088 if $tmp != $ERROR_WANT_READ
2089 && ($tmp != $ERROR_SYSCALL || $!);
2090
2091 last;
1145 } 2092 }
1146 }
1147 2093
2094 substr $self->{_tls_wbuf}, 0, $tmp, "";
2095 }
2096
2097 while (defined ($tmp = Net::SSLeay::read ($self->{tls}))) {
2098 unless (length $tmp) {
2099 $self->{_on_starttls}
2100 and (delete $self->{_on_starttls})->($self, undef, "EOF during handshake"); # ???
2101 &_freetls;
2102
2103 if ($self->{on_stoptls}) {
2104 $self->{on_stoptls}($self);
2105 return;
2106 } else {
2107 # let's treat SSL-eof as we treat normal EOF
2108 delete $self->{_rw};
2109 $self->{_eof} = 1;
2110 }
2111 }
2112
2113 $self->{_tls_rbuf} .= $tmp;
2114 $self->_drain_rbuf;
2115 $self->{tls} or return; # tls session might have gone away in callback
2116 }
2117
2118 $tmp = Net::SSLeay::get_error ($self->{tls}, -1); # -1 is not neccessarily correct, but Net::SSLeay doesn't tell us
2119 return $self->_tls_error ($tmp)
2120 if $tmp != $ERROR_WANT_READ
2121 && ($tmp != $ERROR_SYSCALL || $!);
2122
1148 if (length ($buf = Net::SSLeay::BIO_read ($self->{_wbio}))) { 2123 while (length ($tmp = Net::SSLeay::BIO_read ($self->{_wbio}))) {
1149 $self->{wbuf} .= $buf; 2124 $self->{wbuf} .= $tmp;
1150 $self->_drain_wbuf; 2125 $self->_drain_wbuf;
2126 $self->{tls} or return; # tls session might have gone away in callback
1151 } 2127 }
1152 2128
1153 while (defined ($buf = Net::SSLeay::read ($self->{tls}))) { 2129 $self->{_on_starttls}
1154 if (length $buf) { 2130 and Net::SSLeay::state ($self->{tls}) == Net::SSLeay::ST_OK ()
1155 $self->{rbuf} .= $buf; 2131 and (delete $self->{_on_starttls})->($self, 1, "TLS/SSL connection established");
1156 $self->_drain_rbuf unless $self->{_in_drain};
1157 } else {
1158 # let's treat SSL-eof as we treat normal EOF
1159 $self->{_eof} = 1;
1160 $self->_shutdown;
1161 return;
1162 }
1163 }
1164
1165 my $err = Net::SSLeay::get_error ($self->{tls}, -1);
1166
1167 if ($err!= Net::SSLeay::ERROR_WANT_READ ()) {
1168 if ($err == Net::SSLeay::ERROR_SYSCALL ()) {
1169 return $self->_error ($!, 1);
1170 } elsif ($err == Net::SSLeay::ERROR_SSL ()) {
1171 return $self->_error (&Errno::EIO, 1);
1172 }
1173
1174 # all others are fine for our purposes
1175 }
1176} 2132}
1177 2133
1178=item $handle->starttls ($tls[, $tls_ctx]) 2134=item $handle->starttls ($tls[, $tls_ctx])
1179 2135
1180Instead of starting TLS negotiation immediately when the AnyEvent::Handle 2136Instead of starting TLS negotiation immediately when the AnyEvent::Handle
1181object is created, you can also do that at a later time by calling 2137object is created, you can also do that at a later time by calling
1182C<starttls>. 2138C<starttls>. See the C<tls> constructor argument for general info.
2139
2140Starting TLS is currently an asynchronous operation - when you push some
2141write data and then call C<< ->starttls >> then TLS negotiation will start
2142immediately, after which the queued write data is then sent. This might
2143change in future versions, so best make sure you have no outstanding write
2144data when calling this method.
1183 2145
1184The first argument is the same as the C<tls> constructor argument (either 2146The first argument is the same as the C<tls> constructor argument (either
1185C<"connect">, C<"accept"> or an existing Net::SSLeay object). 2147C<"connect">, C<"accept"> or an existing Net::SSLeay object).
1186 2148
1187The second argument is the optional C<Net::SSLeay::CTX> object that is 2149The second argument is the optional C<AnyEvent::TLS> object that is used
1188used when AnyEvent::Handle has to create its own TLS connection object. 2150when AnyEvent::Handle has to create its own TLS connection object, or
2151a hash reference with C<< key => value >> pairs that will be used to
2152construct a new context.
1189 2153
1190The TLS connection object will end up in C<< $handle->{tls} >> after this 2154The TLS connection object will end up in C<< $handle->{tls} >>, the TLS
1191call and can be used or changed to your liking. Note that the handshake 2155context in C<< $handle->{tls_ctx} >> after this call and can be used or
1192might have already started when this function returns. 2156changed to your liking. Note that the handshake might have already started
2157when this function returns.
1193 2158
2159Due to bugs in OpenSSL, it might or might not be possible to do multiple
2160handshakes on the same stream. It is best to not attempt to use the
2161stream after stopping TLS.
2162
2163This method may invoke callbacks (and therefore the handle might be
2164destroyed after it returns).
2165
1194=cut 2166=cut
2167
2168our %TLS_CACHE; #TODO not yet documented, should we?
1195 2169
1196sub starttls { 2170sub starttls {
1197 my ($self, $ssl, $ctx) = @_; 2171 my ($self, $tls, $ctx) = @_;
1198 2172
1199 $self->stoptls; 2173 Carp::croak "It is an error to call starttls on an AnyEvent::Handle object while TLS is already active, caught"
2174 if $self->{tls};
1200 2175
1201 if ($ssl eq "accept") { 2176 unless (defined $AnyEvent::TLS::VERSION) {
1202 $ssl = Net::SSLeay::new ($ctx || TLS_CTX ()); 2177 eval {
1203 Net::SSLeay::set_accept_state ($ssl); 2178 require Net::SSLeay;
1204 } elsif ($ssl eq "connect") { 2179 require AnyEvent::TLS;
1205 $ssl = Net::SSLeay::new ($ctx || TLS_CTX ()); 2180 1
1206 Net::SSLeay::set_connect_state ($ssl); 2181 } or return $self->_error (Errno::EPROTO, 1, "TLS support not available on this system");
2182 }
2183
2184 $self->{tls} = $tls;
2185 $self->{tls_ctx} = $ctx if @_ > 2;
2186
2187 return unless $self->{fh};
2188
2189 $ERROR_SYSCALL = Net::SSLeay::ERROR_SYSCALL ();
2190 $ERROR_WANT_READ = Net::SSLeay::ERROR_WANT_READ ();
2191
2192 $tls = delete $self->{tls};
2193 $ctx = $self->{tls_ctx};
2194
2195 local $Carp::CarpLevel = 1; # skip ourselves when creating a new context or session
2196
2197 if ("HASH" eq ref $ctx) {
2198 if ($ctx->{cache}) {
2199 my $key = $ctx+0;
2200 $ctx = $TLS_CACHE{$key} ||= new AnyEvent::TLS %$ctx;
2201 } else {
2202 $ctx = new AnyEvent::TLS %$ctx;
2203 }
2204 }
1207 } 2205
1208 2206 $self->{tls_ctx} = $ctx || TLS_CTX ();
1209 $self->{tls} = $ssl; 2207 $self->{tls} = $tls = $self->{tls_ctx}->_get_session ($tls, $self, $self->{peername});
1210 2208
1211 # basically, this is deep magic (because SSL_read should have the same issues) 2209 # basically, this is deep magic (because SSL_read should have the same issues)
1212 # but the openssl maintainers basically said: "trust us, it just works". 2210 # but the openssl maintainers basically said: "trust us, it just works".
1213 # (unfortunately, we have to hardcode constants because the abysmally misdesigned 2211 # (unfortunately, we have to hardcode constants because the abysmally misdesigned
1214 # and mismaintained ssleay-module doesn't even offer them). 2212 # and mismaintained ssleay-module doesn't even offer them).
1215 # http://www.mail-archive.com/openssl-dev@openssl.org/msg22420.html 2213 # http://www.mail-archive.com/openssl-dev@openssl.org/msg22420.html
2214 #
2215 # in short: this is a mess.
2216 #
2217 # note that we do not try to keep the length constant between writes as we are required to do.
2218 # we assume that most (but not all) of this insanity only applies to non-blocking cases,
2219 # and we drive openssl fully in blocking mode here. Or maybe we don't - openssl seems to
2220 # have identity issues in that area.
1216 Net::SSLeay::CTX_set_mode ($self->{tls}, 2221# Net::SSLeay::CTX_set_mode ($ssl,
1217 (eval { local $SIG{__DIE__}; Net::SSLeay::MODE_ENABLE_PARTIAL_WRITE () } || 1) 2222# (eval { local $SIG{__DIE__}; Net::SSLeay::MODE_ENABLE_PARTIAL_WRITE () } || 1)
1218 | (eval { local $SIG{__DIE__}; Net::SSLeay::MODE_ACCEPT_MOVING_WRITE_BUFFER () } || 2)); 2223# | (eval { local $SIG{__DIE__}; Net::SSLeay::MODE_ACCEPT_MOVING_WRITE_BUFFER () } || 2));
2224 Net::SSLeay::CTX_set_mode ($tls, 1|2);
1219 2225
1220 $self->{_rbio} = Net::SSLeay::BIO_new (Net::SSLeay::BIO_s_mem ()); 2226 $self->{_rbio} = Net::SSLeay::BIO_new (Net::SSLeay::BIO_s_mem ());
1221 $self->{_wbio} = Net::SSLeay::BIO_new (Net::SSLeay::BIO_s_mem ()); 2227 $self->{_wbio} = Net::SSLeay::BIO_new (Net::SSLeay::BIO_s_mem ());
1222 2228
2229 Net::SSLeay::BIO_write ($self->{_rbio}, $self->{rbuf});
2230 $self->{rbuf} = "";
2231
1223 Net::SSLeay::set_bio ($ssl, $self->{_rbio}, $self->{_wbio}); 2232 Net::SSLeay::set_bio ($tls, $self->{_rbio}, $self->{_wbio});
1224 2233
1225 $self->{filter_w} = sub { 2234 $self->{_on_starttls} = sub { $_[0]{on_starttls}(@_) }
1226 $_[0]{_tls_wbuf} .= ${$_[1]}; 2235 if $self->{on_starttls};
1227 &_dotls; 2236
1228 }; 2237 &_dotls; # need to trigger the initial handshake
1229 $self->{filter_r} = sub { 2238 $self->start_read; # make sure we actually do read
1230 Net::SSLeay::BIO_write ($_[0]{_rbio}, ${$_[1]});
1231 &_dotls;
1232 };
1233} 2239}
1234 2240
1235=item $handle->stoptls 2241=item $handle->stoptls
1236 2242
1237Destroys the SSL connection, if any. Partial read or write data will be 2243Shuts down the SSL connection - this makes a proper EOF handshake by
1238lost. 2244sending a close notify to the other side, but since OpenSSL doesn't
2245support non-blocking shut downs, it is not guaranteed that you can re-use
2246the stream afterwards.
2247
2248This method may invoke callbacks (and therefore the handle might be
2249destroyed after it returns).
1239 2250
1240=cut 2251=cut
1241 2252
1242sub stoptls { 2253sub stoptls {
1243 my ($self) = @_; 2254 my ($self) = @_;
1244 2255
1245 Net::SSLeay::free (delete $self->{tls}) if $self->{tls}; 2256 if ($self->{tls} && $self->{fh}) {
2257 Net::SSLeay::shutdown ($self->{tls});
1246 2258
1247 delete $self->{_rbio}; 2259 &_dotls;
1248 delete $self->{_wbio}; 2260
1249 delete $self->{_tls_wbuf}; 2261# # we don't give a shit. no, we do, but we can't. no...#d#
1250 delete $self->{filter_r}; 2262# # we, we... have to use openssl :/#d#
1251 delete $self->{filter_w}; 2263# &_freetls;#d#
2264 }
1252} 2265}
2266
2267sub _freetls {
2268 my ($self) = @_;
2269
2270 return unless $self->{tls};
2271
2272 $self->{tls_ctx}->_put_session (delete $self->{tls})
2273 if $self->{tls} > 0;
2274
2275 delete @$self{qw(_rbio _wbio _tls_wbuf _on_starttls)};
2276}
2277
2278=item $handle->resettls
2279
2280This rarely-used method simply resets and TLS state on the handle, usually
2281causing data loss.
2282
2283One case where it may be useful is when you want to skip over the data in
2284the stream but you are not interested in interpreting it, so data loss is
2285no concern.
2286
2287=cut
2288
2289*resettls = \&_freetls;
1253 2290
1254sub DESTROY { 2291sub DESTROY {
1255 my $self = shift; 2292 my ($self) = @_;
1256 2293
1257 $self->stoptls; 2294 &_freetls;
2295
2296 my $linger = exists $self->{linger} ? $self->{linger} : 3600;
2297
2298 if ($linger && length $self->{wbuf} && $self->{fh}) {
2299 my $fh = delete $self->{fh};
2300 my $wbuf = delete $self->{wbuf};
2301
2302 my @linger;
2303
2304 push @linger, AE::io $fh, 1, sub {
2305 my $len = syswrite $fh, $wbuf, length $wbuf;
2306
2307 if ($len > 0) {
2308 substr $wbuf, 0, $len, "";
2309 } elsif (defined $len || ($! != EAGAIN && $! != EINTR && $! != EWOULDBLOCK && $! != WSAEWOULDBLOCK)) {
2310 @linger = (); # end
2311 }
2312 };
2313 push @linger, AE::timer $linger, 0, sub {
2314 @linger = ();
2315 };
2316 }
1258} 2317}
2318
2319=item $handle->destroy
2320
2321Shuts down the handle object as much as possible - this call ensures that
2322no further callbacks will be invoked and as many resources as possible
2323will be freed. Any method you will call on the handle object after
2324destroying it in this way will be silently ignored (and it will return the
2325empty list).
2326
2327Normally, you can just "forget" any references to an AnyEvent::Handle
2328object and it will simply shut down. This works in fatal error and EOF
2329callbacks, as well as code outside. It does I<NOT> work in a read or write
2330callback, so when you want to destroy the AnyEvent::Handle object from
2331within such an callback. You I<MUST> call C<< ->destroy >> explicitly in
2332that case.
2333
2334Destroying the handle object in this way has the advantage that callbacks
2335will be removed as well, so if those are the only reference holders (as
2336is common), then one doesn't need to do anything special to break any
2337reference cycles.
2338
2339The handle might still linger in the background and write out remaining
2340data, as specified by the C<linger> option, however.
2341
2342=cut
2343
2344sub destroy {
2345 my ($self) = @_;
2346
2347 $self->DESTROY;
2348 %$self = ();
2349 bless $self, "AnyEvent::Handle::destroyed";
2350}
2351
2352sub AnyEvent::Handle::destroyed::AUTOLOAD {
2353 #nop
2354}
2355
2356=item $handle->destroyed
2357
2358Returns false as long as the handle hasn't been destroyed by a call to C<<
2359->destroy >>, true otherwise.
2360
2361Can be useful to decide whether the handle is still valid after some
2362callback possibly destroyed the handle. For example, C<< ->push_write >>,
2363C<< ->starttls >> and other methods can call user callbacks, which in turn
2364can destroy the handle, so work can be avoided by checking sometimes:
2365
2366 $hdl->starttls ("accept");
2367 return if $hdl->destroyed;
2368 $hdl->push_write (...
2369
2370Note that the call to C<push_write> will silently be ignored if the handle
2371has been destroyed, so often you can just ignore the possibility of the
2372handle being destroyed.
2373
2374=cut
2375
2376sub destroyed { 0 }
2377sub AnyEvent::Handle::destroyed::destroyed { 1 }
1259 2378
1260=item AnyEvent::Handle::TLS_CTX 2379=item AnyEvent::Handle::TLS_CTX
1261 2380
1262This function creates and returns the Net::SSLeay::CTX object used by 2381This function creates and returns the AnyEvent::TLS object used by default
1263default for TLS mode. 2382for TLS mode.
1264 2383
1265The context is created like this: 2384The context is created by calling L<AnyEvent::TLS> without any arguments.
1266
1267 Net::SSLeay::load_error_strings;
1268 Net::SSLeay::SSLeay_add_ssl_algorithms;
1269 Net::SSLeay::randomize;
1270
1271 my $CTX = Net::SSLeay::CTX_new;
1272
1273 Net::SSLeay::CTX_set_options $CTX, Net::SSLeay::OP_ALL
1274 2385
1275=cut 2386=cut
1276 2387
1277our $TLS_CTX; 2388our $TLS_CTX;
1278 2389
1279sub TLS_CTX() { 2390sub TLS_CTX() {
1280 $TLS_CTX || do { 2391 $TLS_CTX ||= do {
1281 require Net::SSLeay; 2392 require AnyEvent::TLS;
1282 2393
1283 Net::SSLeay::load_error_strings (); 2394 new AnyEvent::TLS
1284 Net::SSLeay::SSLeay_add_ssl_algorithms (); 2395 }
1285 Net::SSLeay::randomize (); 2396}
1286 2397
1287 $TLS_CTX = Net::SSLeay::CTX_new (); 2398=back
1288 2399
1289 Net::SSLeay::CTX_set_options ($TLS_CTX, Net::SSLeay::OP_ALL ());
1290 2400
1291 $TLS_CTX 2401=head1 NONFREQUENTLY ASKED QUESTIONS
2402
2403=over 4
2404
2405=item I C<undef> the AnyEvent::Handle reference inside my callback and
2406still get further invocations!
2407
2408That's because AnyEvent::Handle keeps a reference to itself when handling
2409read or write callbacks.
2410
2411It is only safe to "forget" the reference inside EOF or error callbacks,
2412from within all other callbacks, you need to explicitly call the C<<
2413->destroy >> method.
2414
2415=item Why is my C<on_eof> callback never called?
2416
2417Probably because your C<on_error> callback is being called instead: When
2418you have outstanding requests in your read queue, then an EOF is
2419considered an error as you clearly expected some data.
2420
2421To avoid this, make sure you have an empty read queue whenever your handle
2422is supposed to be "idle" (i.e. connection closes are O.K.). You can set
2423an C<on_read> handler that simply pushes the first read requests in the
2424queue.
2425
2426See also the next question, which explains this in a bit more detail.
2427
2428=item How can I serve requests in a loop?
2429
2430Most protocols consist of some setup phase (authentication for example)
2431followed by a request handling phase, where the server waits for requests
2432and handles them, in a loop.
2433
2434There are two important variants: The first (traditional, better) variant
2435handles requests until the server gets some QUIT command, causing it to
2436close the connection first (highly desirable for a busy TCP server). A
2437client dropping the connection is an error, which means this variant can
2438detect an unexpected detection close.
2439
2440To handle this case, always make sure you have a non-empty read queue, by
2441pushing the "read request start" handler on it:
2442
2443 # we assume a request starts with a single line
2444 my @start_request; @start_request = (line => sub {
2445 my ($hdl, $line) = @_;
2446
2447 ... handle request
2448
2449 # push next request read, possibly from a nested callback
2450 $hdl->push_read (@start_request);
2451 });
2452
2453 # auth done, now go into request handling loop
2454 # now push the first @start_request
2455 $hdl->push_read (@start_request);
2456
2457By always having an outstanding C<push_read>, the handle always expects
2458some data and raises the C<EPIPE> error when the connction is dropped
2459unexpectedly.
2460
2461The second variant is a protocol where the client can drop the connection
2462at any time. For TCP, this means that the server machine may run out of
2463sockets easier, and in general, it means you cannot distinguish a protocl
2464failure/client crash from a normal connection close. Nevertheless, these
2465kinds of protocols are common (and sometimes even the best solution to the
2466problem).
2467
2468Having an outstanding read request at all times is possible if you ignore
2469C<EPIPE> errors, but this doesn't help with when the client drops the
2470connection during a request, which would still be an error.
2471
2472A better solution is to push the initial request read in an C<on_read>
2473callback. This avoids an error, as when the server doesn't expect data
2474(i.e. is idly waiting for the next request, an EOF will not raise an
2475error, but simply result in an C<on_eof> callback. It is also a bit slower
2476and simpler:
2477
2478 # auth done, now go into request handling loop
2479 $hdl->on_read (sub {
2480 my ($hdl) = @_;
2481
2482 # called each time we receive data but the read queue is empty
2483 # simply start read the request
2484
2485 $hdl->push_read (line => sub {
2486 my ($hdl, $line) = @_;
2487
2488 ... handle request
2489
2490 # do nothing special when the request has been handled, just
2491 # let the request queue go empty.
2492 });
2493 });
2494
2495=item I get different callback invocations in TLS mode/Why can't I pause
2496reading?
2497
2498Unlike, say, TCP, TLS connections do not consist of two independent
2499communication channels, one for each direction. Or put differently, the
2500read and write directions are not independent of each other: you cannot
2501write data unless you are also prepared to read, and vice versa.
2502
2503This means that, in TLS mode, you might get C<on_error> or C<on_eof>
2504callback invocations when you are not expecting any read data - the reason
2505is that AnyEvent::Handle always reads in TLS mode.
2506
2507During the connection, you have to make sure that you always have a
2508non-empty read-queue, or an C<on_read> watcher. At the end of the
2509connection (or when you no longer want to use it) you can call the
2510C<destroy> method.
2511
2512=item How do I read data until the other side closes the connection?
2513
2514If you just want to read your data into a perl scalar, the easiest way
2515to achieve this is by setting an C<on_read> callback that does nothing,
2516clearing the C<on_eof> callback and in the C<on_error> callback, the data
2517will be in C<$_[0]{rbuf}>:
2518
2519 $handle->on_read (sub { });
2520 $handle->on_eof (undef);
2521 $handle->on_error (sub {
2522 my $data = delete $_[0]{rbuf};
2523 });
2524
2525Note that this example removes the C<rbuf> member from the handle object,
2526which is not normally allowed by the API. It is expressly permitted in
2527this case only, as the handle object needs to be destroyed afterwards.
2528
2529The reason to use C<on_error> is that TCP connections, due to latencies
2530and packets loss, might get closed quite violently with an error, when in
2531fact all data has been received.
2532
2533It is usually better to use acknowledgements when transferring data,
2534to make sure the other side hasn't just died and you got the data
2535intact. This is also one reason why so many internet protocols have an
2536explicit QUIT command.
2537
2538=item I don't want to destroy the handle too early - how do I wait until
2539all data has been written?
2540
2541After writing your last bits of data, set the C<on_drain> callback
2542and destroy the handle in there - with the default setting of
2543C<low_water_mark> this will be called precisely when all data has been
2544written to the socket:
2545
2546 $handle->push_write (...);
2547 $handle->on_drain (sub {
2548 AE::log debug => "All data submitted to the kernel.";
2549 undef $handle;
2550 });
2551
2552If you just want to queue some data and then signal EOF to the other side,
2553consider using C<< ->push_shutdown >> instead.
2554
2555=item I want to contact a TLS/SSL server, I don't care about security.
2556
2557If your TLS server is a pure TLS server (e.g. HTTPS) that only speaks TLS,
2558connect to it and then create the AnyEvent::Handle with the C<tls>
2559parameter:
2560
2561 tcp_connect $host, $port, sub {
2562 my ($fh) = @_;
2563
2564 my $handle = new AnyEvent::Handle
2565 fh => $fh,
2566 tls => "connect",
2567 on_error => sub { ... };
2568
2569 $handle->push_write (...);
1292 } 2570 };
1293} 2571
2572=item I want to contact a TLS/SSL server, I do care about security.
2573
2574Then you should additionally enable certificate verification, including
2575peername verification, if the protocol you use supports it (see
2576L<AnyEvent::TLS>, C<verify_peername>).
2577
2578E.g. for HTTPS:
2579
2580 tcp_connect $host, $port, sub {
2581 my ($fh) = @_;
2582
2583 my $handle = new AnyEvent::Handle
2584 fh => $fh,
2585 peername => $host,
2586 tls => "connect",
2587 tls_ctx => { verify => 1, verify_peername => "https" },
2588 ...
2589
2590Note that you must specify the hostname you connected to (or whatever
2591"peername" the protocol needs) as the C<peername> argument, otherwise no
2592peername verification will be done.
2593
2594The above will use the system-dependent default set of trusted CA
2595certificates. If you want to check against a specific CA, add the
2596C<ca_file> (or C<ca_cert>) arguments to C<tls_ctx>:
2597
2598 tls_ctx => {
2599 verify => 1,
2600 verify_peername => "https",
2601 ca_file => "my-ca-cert.pem",
2602 },
2603
2604=item I want to create a TLS/SSL server, how do I do that?
2605
2606Well, you first need to get a server certificate and key. You have
2607three options: a) ask a CA (buy one, use cacert.org etc.) b) create a
2608self-signed certificate (cheap. check the search engine of your choice,
2609there are many tutorials on the net) or c) make your own CA (tinyca2 is a
2610nice program for that purpose).
2611
2612Then create a file with your private key (in PEM format, see
2613L<AnyEvent::TLS>), followed by the certificate (also in PEM format). The
2614file should then look like this:
2615
2616 -----BEGIN RSA PRIVATE KEY-----
2617 ...header data
2618 ... lots of base64'y-stuff
2619 -----END RSA PRIVATE KEY-----
2620
2621 -----BEGIN CERTIFICATE-----
2622 ... lots of base64'y-stuff
2623 -----END CERTIFICATE-----
2624
2625The important bits are the "PRIVATE KEY" and "CERTIFICATE" parts. Then
2626specify this file as C<cert_file>:
2627
2628 tcp_server undef, $port, sub {
2629 my ($fh) = @_;
2630
2631 my $handle = new AnyEvent::Handle
2632 fh => $fh,
2633 tls => "accept",
2634 tls_ctx => { cert_file => "my-server-keycert.pem" },
2635 ...
2636
2637When you have intermediate CA certificates that your clients might not
2638know about, just append them to the C<cert_file>.
1294 2639
1295=back 2640=back
1296 2641
1297=head1 SUBCLASSING AnyEvent::Handle 2642=head1 SUBCLASSING AnyEvent::Handle
1298 2643
1304=over 4 2649=over 4
1305 2650
1306=item * all constructor arguments become object members. 2651=item * all constructor arguments become object members.
1307 2652
1308At least initially, when you pass a C<tls>-argument to the constructor it 2653At least initially, when you pass a C<tls>-argument to the constructor it
1309will end up in C<< $handle->{tls} >>. Those members might be changes or 2654will end up in C<< $handle->{tls} >>. Those members might be changed or
1310mutated later on (for example C<tls> will hold the TLS connection object). 2655mutated later on (for example C<tls> will hold the TLS connection object).
1311 2656
1312=item * other object member names are prefixed with an C<_>. 2657=item * other object member names are prefixed with an C<_>.
1313 2658
1314All object members not explicitly documented (internal use) are prefixed 2659All object members not explicitly documented (internal use) are prefixed
1317 2662
1318=item * all members not documented here and not prefixed with an underscore 2663=item * all members not documented here and not prefixed with an underscore
1319are free to use in subclasses. 2664are free to use in subclasses.
1320 2665
1321Of course, new versions of AnyEvent::Handle may introduce more "public" 2666Of course, new versions of AnyEvent::Handle may introduce more "public"
1322member variables, but thats just life, at least it is documented. 2667member variables, but that's just life. At least it is documented.
1323 2668
1324=back 2669=back
1325 2670
1326=head1 AUTHOR 2671=head1 AUTHOR
1327 2672
1328Robin Redeker C<< <elmex at ta-sa.org> >>, Marc Lehmann <schmorp@schmorp.de>. 2673Robin Redeker C<< <elmex at ta-sa.org> >>, Marc Lehmann <schmorp@schmorp.de>.
1329 2674
1330=cut 2675=cut
1331 2676
13321; # End of AnyEvent::Handle 26771
2678

Diff Legend

Removed lines
+ Added lines
< Changed lines
> Changed lines