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 1.48 by root, Thu May 29 00:27:06 2008 UTC vs.
Revision 1.256 by root, Wed Jul 29 15:58:58 2020 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 = '0.04';
20 4
21=head1 SYNOPSIS 5=head1 SYNOPSIS
22 6
23 use AnyEvent; 7 use AnyEvent;
24 use AnyEvent::Handle; 8 use AnyEvent::Handle;
25 9
26 my $cv = AnyEvent->condvar; 10 my $cv = AnyEvent->condvar;
27 11
28 my $handle = 12 my $hdl; $hdl = new AnyEvent::Handle
29 AnyEvent::Handle->new (
30 fh => \*STDIN, 13 fh => \*STDIN,
31 on_eof => sub { 14 on_error => sub {
32 $cv->broadcast; 15 my ($hdl, $fatal, $msg) = @_;
33 }, 16 AE::log error => $msg;
17 $hdl->destroy;
18 $cv->send;
34 ); 19 };
35 20
36 # send some request line 21 # send some request line
37 $handle->push_write ("getinfo\015\012"); 22 $hdl->push_write ("getinfo\015\012");
38 23
39 # read the response line 24 # read the response line
40 $handle->push_read (line => sub { 25 $hdl->push_read (line => sub {
41 my ($handle, $line) = @_; 26 my ($hdl, $line) = @_;
42 warn "read line <$line>\n"; 27 say "got line <$line>";
43 $cv->send; 28 $cv->send;
44 }); 29 });
45 30
46 $cv->recv; 31 $cv->recv;
47 32
48=head1 DESCRIPTION 33=head1 DESCRIPTION
49 34
50This module is a helper module to make it easier to do event-based I/O on 35This is a helper module to make it easier to do event-based I/O
51filehandles. For utility functions for doing non-blocking connects and accepts 36on stream-based filehandles (sockets, pipes, and other stream
52on sockets see L<AnyEvent::Util>. 37things). Specifically, it doesn't work as expected on files, packet-based
38sockets or similar things.
53 39
40The L<AnyEvent::Intro> tutorial contains some well-documented
41AnyEvent::Handle examples.
42
54In the following, when the documentation refers to of "bytes" then this 43In the following, where the documentation refers to "bytes", it means
55means characters. As sysread and syswrite are used for all I/O, their 44characters. As sysread and syswrite are used for all I/O, their
56treatment of characters applies to this module as well. 45treatment of characters applies to this module as well.
46
47At the very minimum, you should specify C<fh> or C<connect>, and the
48C<on_error> callback.
57 49
58All callbacks will be invoked with the handle object as their first 50All callbacks will be invoked with the handle object as their first
59argument. 51argument.
60 52
53=cut
54
55package AnyEvent::Handle;
56
57use Scalar::Util ();
58use List::Util ();
59use Carp ();
60use Errno qw(EAGAIN EWOULDBLOCK EINTR);
61
62use AnyEvent (); BEGIN { AnyEvent::common_sense }
63use AnyEvent::Util qw(WSAEWOULDBLOCK);
64
65our $VERSION = $AnyEvent::VERSION;
66
67sub _load_func($) {
68 my $func = $_[0];
69
70 unless (defined &$func) {
71 my $pkg = $func;
72 do {
73 $pkg =~ s/::[^:]+$//
74 or return;
75 eval "require $pkg";
76 } until defined &$func;
77 }
78
79 \&$func
80}
81
82sub MAX_READ_SIZE() { 131072 }
83
61=head1 METHODS 84=head1 METHODS
62 85
63=over 4 86=over 4
64 87
65=item B<new (%args)> 88=item $handle = B<new> AnyEvent::Handle fh => $filehandle, key => value...
66 89
67The constructor supports these arguments (all as key => value pairs). 90The constructor supports these arguments (all as C<< key => value >> pairs).
68 91
69=over 4 92=over 4
70 93
71=item fh => $filehandle [MANDATORY] 94=item fh => $filehandle [C<fh> or C<connect> MANDATORY]
72 95
73The filehandle this L<AnyEvent::Handle> object will operate on. 96The filehandle this L<AnyEvent::Handle> object will operate on.
74
75NOTE: The filehandle will be set to non-blocking (using 97NOTE: The filehandle will be set to non-blocking mode (using
76AnyEvent::Util::fh_nonblocking). 98C<AnyEvent::fh_unblock>) by the constructor and needs to stay in
99that mode.
77 100
78=item on_eof => $cb->($handle) 101=item connect => [$host, $service] [C<fh> or C<connect> MANDATORY]
79 102
80Set the callback to be called on EOF. 103Try to connect to the specified host and service (port), using
104C<AnyEvent::Socket::tcp_connect>. The C<$host> additionally becomes the
105default C<peername>.
81 106
82While not mandatory, it is highly recommended to set an eof callback, 107You have to specify either this parameter, or C<fh>, above.
83otherwise you might end up with a closed socket while you are still
84waiting for data.
85 108
109It is possible to push requests on the read and write queues, and modify
110properties of the stream, even while AnyEvent::Handle is connecting.
111
112When this parameter is specified, then the C<on_prepare>,
113C<on_connect_error> and C<on_connect> callbacks will be called under the
114appropriate circumstances:
115
116=over 4
117
86=item on_error => $cb->($handle) 118=item on_prepare => $cb->($handle)
87 119
120This (rarely used) callback is called before a new connection is
121attempted, but after the file handle has been created (you can access that
122file handle via C<< $handle->{fh} >>). It could be used to prepare the
123file handle with parameters required for the actual connect (as opposed to
124settings that can be changed when the connection is already established).
125
126The return value of this callback should be the connect timeout value in
127seconds (or C<0>, or C<undef>, or the empty list, to indicate that the
128default timeout is to be used).
129
130=item on_connect => $cb->($handle, $host, $port, $retry->())
131
132This callback is called when a connection has been successfully established.
133
134The peer's numeric host and port (the socket peername) are passed as
135parameters, together with a retry callback. At the time it is called the
136read and write queues, EOF status, TLS status and similar properties of
137the handle will have been reset.
138
139If, for some reason, the handle is not acceptable, calling C<$retry> will
140continue with the next connection target (in case of multi-homed hosts or
141SRV records there can be multiple connection endpoints). The C<$retry>
142callback can be invoked after the connect callback returns, i.e. one can
143start a handshake and then decide to retry with the next host if the
144handshake fails.
145
146In most cases, you should ignore the C<$retry> parameter.
147
148=item on_connect_error => $cb->($handle, $message)
149
150This callback is called when the connection could not be
151established. C<$!> will contain the relevant error code, and C<$message> a
152message describing it (usually the same as C<"$!">).
153
154If this callback isn't specified, then C<on_error> will be called with a
155fatal error instead.
156
157=back
158
159=item on_error => $cb->($handle, $fatal, $message)
160
88This is the fatal error callback, that is called when, well, a fatal error 161This is the error callback, which is called when, well, some error
89occurs, such as not being able to resolve the hostname, failure to connect 162occured, such as not being able to resolve the hostname, failure to
90or a read error. 163connect, or a read error.
91 164
92The object will not be in a usable state when this callback has been 165Some errors are fatal (which is indicated by C<$fatal> being true). On
93called. 166fatal errors the handle object will be destroyed (by a call to C<< ->
167destroy >>) after invoking the error callback (which means you are free to
168examine the handle object). Examples of fatal errors are an EOF condition
169with active (but unsatisfiable) read watchers (C<EPIPE>) or I/O errors. In
170cases where the other side can close the connection at will, it is
171often easiest to not report C<EPIPE> errors in this callback.
94 172
173AnyEvent::Handle tries to find an appropriate error code for you to check
174against, but in some cases (TLS errors), this does not work well.
175
176If you report the error to the user, it is recommended to always output
177the C<$message> argument in human-readable error messages (you don't need
178to report C<"$!"> if you report C<$message>).
179
180If you want to react programmatically to the error, then looking at C<$!>
181and comparing it against some of the documented C<Errno> values is usually
182better than looking at the C<$message>.
183
184Non-fatal errors can be retried by returning, but it is recommended
185to simply ignore this parameter and instead abondon the handle object
186when this callback is invoked. Examples of non-fatal errors are timeouts
187C<ETIMEDOUT>) or badly-formatted data (C<EBADMSG>).
188
95On callback entrance, the value of C<$!> contains the operating system 189On entry to the callback, the value of C<$!> contains the operating
96error (or C<ENOSPC>, C<EPIPE>, C<ETIMEDOUT> or C<EBADMSG>). 190system error code (or C<ENOSPC>, C<EPIPE>, C<ETIMEDOUT>, C<EBADMSG> or
97 191C<EPROTO>).
98The callback should throw an exception. If it returns, then
99AnyEvent::Handle will C<croak> for you.
100 192
101While not mandatory, it is I<highly> recommended to set this callback, as 193While not mandatory, it is I<highly> recommended to set this callback, as
102you will not be notified of errors otherwise. The default simply calls 194you will not be notified of errors otherwise. The default just calls
103die. 195C<croak>.
104 196
105=item on_read => $cb->($handle) 197=item on_read => $cb->($handle)
106 198
107This sets the default read callback, which is called when data arrives 199This sets the default read callback, which is called when data arrives
108and no read request is in the queue. 200and no read request is in the queue (unlike read queue callbacks, this
201callback will only be called when at least one octet of data is in the
202read buffer).
109 203
110To access (and remove data from) the read buffer, use the C<< ->rbuf >> 204To access (and remove data from) the read buffer, use the C<< ->rbuf >>
111method or access the C<$handle->{rbuf}> member directly. 205method or access the C<< $handle->{rbuf} >> member directly. Note that you
206must not enlarge or modify the read buffer, you can only remove data at
207the beginning from it.
112 208
209You can also call C<< ->push_read (...) >> or any other function that
210modifies the read queue. Or do both. Or ...
211
113When an EOF condition is detected then AnyEvent::Handle will first try to 212When an EOF condition is detected, AnyEvent::Handle will first try to
114feed all the remaining data to the queued callbacks and C<on_read> before 213feed all the remaining data to the queued callbacks and C<on_read> before
115calling the C<on_eof> callback. If no progress can be made, then a fatal 214calling the C<on_eof> callback. If no progress can be made, then a fatal
116error will be raised (with C<$!> set to C<EPIPE>). 215error will be raised (with C<$!> set to C<EPIPE>).
117 216
217Note that, unlike requests in the read queue, an C<on_read> callback
218doesn't mean you I<require> some data: if there is an EOF and there
219are outstanding read requests then an error will be flagged. With an
220C<on_read> callback, the C<on_eof> callback will be invoked.
221
222=item on_eof => $cb->($handle)
223
224Set the callback to be called when an end-of-file condition is detected,
225i.e. in the case of a socket, when the other side has closed the
226connection cleanly, and there are no outstanding read requests in the
227queue (if there are read requests, then an EOF counts as an unexpected
228connection close and will be flagged as an error).
229
230For sockets, this just means that the other side has stopped sending data,
231you can still try to write data, and, in fact, one can return from the EOF
232callback and continue writing data, as only the read part has been shut
233down.
234
235If an EOF condition has been detected but no C<on_eof> callback has been
236set, then a fatal error will be raised with C<$!> set to <0>.
237
118=item on_drain => $cb->($handle) 238=item on_drain => $cb->($handle)
119 239
120This sets the callback that is called when the write buffer becomes empty 240This sets the callback that is called once when the write buffer becomes
121(or when the callback is set and the buffer is empty already). 241empty (and immediately when the handle object is created).
122 242
123To append to the write buffer, use the C<< ->push_write >> method. 243To append to the write buffer, use the C<< ->push_write >> method.
124 244
245This callback is useful when you don't want to put all of your write data
246into the queue at once, for example, when you want to write the contents
247of some file to the socket you might not want to read the whole file into
248memory and push it into the queue, but instead only read more data from
249the file when the write queue becomes empty.
250
125=item timeout => $fractional_seconds 251=item timeout => $fractional_seconds
126 252
253=item rtimeout => $fractional_seconds
254
255=item wtimeout => $fractional_seconds
256
127If non-zero, then this enables an "inactivity" timeout: whenever this many 257If non-zero, then these enables an "inactivity" timeout: whenever this
128seconds pass without a successful read or write on the underlying file 258many seconds pass without a successful read or write on the underlying
129handle, the C<on_timeout> callback will be invoked (and if that one is 259file handle (or a call to C<timeout_reset>), the C<on_timeout> callback
130missing, an C<ETIMEDOUT> error will be raised). 260will be invoked (and if that one is missing, a non-fatal C<ETIMEDOUT>
261error will be raised).
131 262
263There are three variants of the timeouts that work independently of each
264other, for both read and write (triggered when nothing was read I<OR>
265written), just read (triggered when nothing was read), and just write:
266C<timeout>, C<rtimeout> and C<wtimeout>, with corresponding callbacks
267C<on_timeout>, C<on_rtimeout> and C<on_wtimeout>, and reset functions
268C<timeout_reset>, C<rtimeout_reset>, and C<wtimeout_reset>.
269
132Note that timeout processing is also active when you currently do not have 270Note that timeout processing is active even when you do not have any
133any outstanding read or write requests: If you plan to keep the connection 271outstanding read or write requests: If you plan to keep the connection
134idle then you should disable the timout temporarily or ignore the timeout 272idle then you should disable the timeout temporarily or ignore the
135in the C<on_timeout> callback. 273timeout in the corresponding C<on_timeout> callback, in which case
274AnyEvent::Handle will simply restart the timeout.
136 275
137Zero (the default) disables this timeout. 276Zero (the default) disables the corresponding timeout.
138 277
139=item on_timeout => $cb->($handle) 278=item on_timeout => $cb->($handle)
279
280=item on_rtimeout => $cb->($handle)
281
282=item on_wtimeout => $cb->($handle)
140 283
141Called whenever the inactivity timeout passes. If you return from this 284Called whenever the inactivity timeout passes. If you return from this
142callback, then the timeout will be reset as if some activity had happened, 285callback, then the timeout will be reset as if some activity had happened,
143so this condition is not fatal in any way. 286so this condition is not fatal in any way.
144 287
145=item rbuf_max => <bytes> 288=item rbuf_max => <bytes>
146 289
147If defined, then a fatal error will be raised (with C<$!> set to C<ENOSPC>) 290If defined, then a fatal error will be raised (with C<$!> set to C<ENOSPC>)
148when the read buffer ever (strictly) exceeds this size. This is useful to 291when the read buffer ever (strictly) exceeds this size. This is useful to
149avoid denial-of-service attacks. 292avoid some forms of denial-of-service attacks.
150 293
151For example, a server accepting connections from untrusted sources should 294For example, a server accepting connections from untrusted sources should
152be configured to accept only so-and-so much data that it cannot act on 295be configured to accept only so-and-so much data that it cannot act on
153(for example, when expecting a line, an attacker could send an unlimited 296(for example, when expecting a line, an attacker could send an unlimited
154amount of data without a callback ever being called as long as the line 297amount of data without a callback ever being called as long as the line
155isn't finished). 298isn't finished).
156 299
300=item wbuf_max => <bytes>
301
302If defined, then a fatal error will be raised (with C<$!> set to C<ENOSPC>)
303when the write buffer ever (strictly) exceeds this size. This is useful to
304avoid some forms of denial-of-service attacks.
305
306Although the units of this parameter is bytes, this is the I<raw> number
307of bytes not yet accepted by the kernel. This can make a difference when
308you e.g. use TLS, as TLS typically makes your write data larger (but it
309can also make it smaller due to compression).
310
311As an example of when this limit is useful, take a chat server that sends
312chat messages to a client. If the client does not read those in a timely
313manner then the send buffer in the server would grow unbounded.
314
315=item autocork => <boolean>
316
317When disabled (the default), C<push_write> will try to immediately
318write the data to the handle if possible. This avoids having to register
319a write watcher and wait for the next event loop iteration, but can
320be inefficient if you write multiple small chunks (on the wire, this
321disadvantage is usually avoided by your kernel's nagle algorithm, see
322C<no_delay>, but this option can save costly syscalls).
323
324When enabled, writes will always be queued till the next event loop
325iteration. This is efficient when you do many small writes per iteration,
326but less efficient when you do a single write only per iteration (or when
327the write buffer often is full). It also increases write latency.
328
329=item no_delay => <boolean>
330
331When doing small writes on sockets, your operating system kernel might
332wait a bit for more data before actually sending it out. This is called
333the Nagle algorithm, and usually it is beneficial.
334
335In some situations you want as low a delay as possible, which can be
336accomplishd by setting this option to a true value.
337
338The default is your operating system's default behaviour (most likely
339enabled). This option explicitly enables or disables it, if possible.
340
341=item keepalive => <boolean>
342
343Enables (default disable) the SO_KEEPALIVE option on the stream socket:
344normally, TCP connections have no time-out once established, so TCP
345connections, once established, can stay alive forever even when the other
346side has long gone. TCP keepalives are a cheap way to take down long-lived
347TCP connections when the other side becomes unreachable. While the default
348is OS-dependent, TCP keepalives usually kick in after around two hours,
349and, if the other side doesn't reply, take down the TCP connection some 10
350to 15 minutes later.
351
352It is harmless to specify this option for file handles that do not support
353keepalives, and enabling it on connections that are potentially long-lived
354is usually a good idea.
355
356=item oobinline => <boolean>
357
358BSD majorly fucked up the implementation of TCP urgent data. The result
359is that almost no OS implements TCP according to the specs, and every OS
360implements it slightly differently.
361
362If you want to handle TCP urgent data, then setting this flag (the default
363is enabled) gives you the most portable way of getting urgent data, by
364putting it into the stream.
365
366Since BSD emulation of OOB data on top of TCP's urgent data can have
367security implications, AnyEvent::Handle sets this flag automatically
368unless explicitly specified. Note that setting this flag after
369establishing a connection I<may> be a bit too late (data loss could
370already have occured on BSD systems), but at least it will protect you
371from most attacks.
372
157=item read_size => <bytes> 373=item read_size => <bytes>
158 374
159The default read block size (the amount of bytes this module will try to read 375The initial read block size, the number of bytes this module will try
160during each (loop iteration). Default: C<8192>. 376to read during each loop iteration. Each handle object will consume
377at least this amount of memory for the read buffer as well, so when
378handling many connections watch out for memory requirements). See also
379C<max_read_size>. Default: C<2048>.
380
381=item max_read_size => <bytes>
382
383The maximum read buffer size used by the dynamic adjustment
384algorithm: Each time AnyEvent::Handle can read C<read_size> bytes in
385one go it will double C<read_size> up to the maximum given by this
386option. Default: C<131072> or C<read_size>, whichever is higher.
161 387
162=item low_water_mark => <bytes> 388=item low_water_mark => <bytes>
163 389
164Sets the amount of bytes (default: C<0>) that make up an "empty" write 390Sets the number of bytes (default: C<0>) that make up an "empty" write
165buffer: If the write reaches this size or gets even samller it is 391buffer: If the buffer reaches this size or gets even samller it is
166considered empty. 392considered empty.
167 393
394Sometimes it can be beneficial (for performance reasons) to add data to
395the write buffer before it is fully drained, but this is a rare case, as
396the operating system kernel usually buffers data as well, so the default
397is good in almost all cases.
398
399=item linger => <seconds>
400
401If this is non-zero (default: C<3600>), the destructor of the
402AnyEvent::Handle object will check whether there is still outstanding
403write data and will install a watcher that will write this data to the
404socket. No errors will be reported (this mostly matches how the operating
405system treats outstanding data at socket close time).
406
407This will not work for partial TLS data that could not be encoded
408yet. This data will be lost. Calling the C<stoptls> method in time might
409help.
410
411=item peername => $string
412
413A string used to identify the remote site - usually the DNS hostname
414(I<not> IDN!) used to create the connection, rarely the IP address.
415
416Apart from being useful in error messages, this string is also used in TLS
417peername verification (see C<verify_peername> in L<AnyEvent::TLS>). This
418verification will be skipped when C<peername> is not specified or is
419C<undef>.
420
168=item tls => "accept" | "connect" | Net::SSLeay::SSL object 421=item tls => "accept" | "connect" | Net::SSLeay::SSL object
169 422
170When this parameter is given, it enables TLS (SSL) mode, that means it 423When this parameter is given, it enables TLS (SSL) mode, that means
171will start making tls handshake and will transparently encrypt/decrypt 424AnyEvent will start a TLS handshake as soon as the connection has been
172data. 425established and will transparently encrypt/decrypt data afterwards.
426
427All TLS protocol errors will be signalled as C<EPROTO>, with an
428appropriate error message.
173 429
174TLS mode requires Net::SSLeay to be installed (it will be loaded 430TLS mode requires Net::SSLeay to be installed (it will be loaded
175automatically when you try to create a TLS handle). 431automatically when you try to create a TLS handle): this module doesn't
432have a dependency on that module, so if your module requires it, you have
433to add the dependency yourself. If Net::SSLeay cannot be loaded or is too
434old, you get an C<EPROTO> error.
176 435
177For the TLS server side, use C<accept>, and for the TLS client side of a 436Unlike TCP, TLS has a server and client side: for the TLS server side, use
178connection, use C<connect> mode. 437C<accept>, and for the TLS client side of a connection, use C<connect>
438mode.
179 439
180You can also provide your own TLS connection object, but you have 440You can also provide your own TLS connection object, but you have
181to make sure that you call either C<Net::SSLeay::set_connect_state> 441to make sure that you call either C<Net::SSLeay::set_connect_state>
182or C<Net::SSLeay::set_accept_state> on it before you pass it to 442or C<Net::SSLeay::set_accept_state> on it before you pass it to
183AnyEvent::Handle. 443AnyEvent::Handle. Also, this module will take ownership of this connection
444object.
184 445
446At some future point, AnyEvent::Handle might switch to another TLS
447implementation, then the option to use your own session object will go
448away.
449
450B<IMPORTANT:> since Net::SSLeay "objects" are really only integers,
451passing in the wrong integer will lead to certain crash. This most often
452happens when one uses a stylish C<< tls => 1 >> and is surprised about the
453segmentation fault.
454
185See the C<starttls> method if you need to start TLs negotiation later. 455Use the C<< ->starttls >> method if you need to start TLS negotiation later.
186 456
187=item tls_ctx => $ssl_ctx 457=item tls_ctx => $anyevent_tls
188 458
189Use the given Net::SSLeay::CTX object to create the new TLS connection 459Use the given C<AnyEvent::TLS> object to create the new TLS connection
190(unless a connection object was specified directly). If this parameter is 460(unless a connection object was specified directly). If this
191missing, then AnyEvent::Handle will use C<AnyEvent::Handle::TLS_CTX>. 461parameter is missing (or C<undef>), then AnyEvent::Handle will use
462C<AnyEvent::Handle::TLS_CTX>.
192 463
464Instead of an object, you can also specify a hash reference with C<< key
465=> value >> pairs. Those will be passed to L<AnyEvent::TLS> to create a
466new TLS context object.
467
468=item on_starttls => $cb->($handle, $success[, $error_message])
469
470This callback will be invoked when the TLS/SSL handshake has finished. If
471C<$success> is true, then the TLS handshake succeeded, otherwise it failed
472(C<on_stoptls> will not be called in this case).
473
474The session in C<< $handle->{tls} >> can still be examined in this
475callback, even when the handshake was not successful.
476
477TLS handshake failures will not cause C<on_error> to be invoked when this
478callback is in effect, instead, the error message will be passed to C<on_starttls>.
479
480Without this callback, handshake failures lead to C<on_error> being
481called as usual.
482
483Note that you cannot just call C<starttls> again in this callback. If you
484need to do that, start an zero-second timer instead whose callback can
485then call C<< ->starttls >> again.
486
487=item on_stoptls => $cb->($handle)
488
489When a SSLv3/TLS shutdown/close notify/EOF is detected and this callback is
490set, then it will be invoked after freeing the TLS session. If it is not,
491then a TLS shutdown condition will be treated like a normal EOF condition
492on the handle.
493
494The session in C<< $handle->{tls} >> can still be examined in this
495callback.
496
497This callback will only be called on TLS shutdowns, not when the
498underlying handle signals EOF.
499
193=item json => JSON or JSON::XS object 500=item json => L<JSON>, L<JSON::PP> or L<JSON::XS> object
194 501
195This is the json coder object used by the C<json> read and write types. 502This is the json coder object used by the C<json> read and write types.
196 503
197If you don't supply it, then AnyEvent::Handle will create and use a 504If you don't supply it, then AnyEvent::Handle will create and use a
198suitable one, which will write and expect UTF-8 encoded JSON texts. 505suitable one (on demand), which will write and expect UTF-8 encoded
506JSON texts (either using L<JSON::XS> or L<JSON>). The written texts are
507guaranteed not to contain any newline character.
199 508
509For security reasons, this encoder will likely I<not> handle numbers and
510strings, only arrays and objects/hashes. The reason is that originally
511JSON was self-delimited, but Dougles Crockford thought it was a splendid
512idea to redefine JSON incompatibly, so this is no longer true.
513
514For protocols that used back-to-back JSON texts, this might lead to
515run-ins, where two or more JSON texts will be interpreted as one JSON
516text.
517
518For this reason, if the default encoder uses L<JSON::XS>, it will default
519to not allowing anything but arrays and objects/hashes, at least for the
520forseeable future (it will change at some point). This might or might not
521be true for the L<JSON> module, so this might cause a security issue.
522
523If you depend on either behaviour, you should create your own json object
524and pass it in explicitly.
525
526=item cbor => L<CBOR::XS> object
527
528This is the cbor coder object used by the C<cbor> read and write types.
529
530If you don't supply it, then AnyEvent::Handle will create and use a
531suitable one (on demand), which will write CBOR without using extensions,
532if possible.
533
200Note that you are responsible to depend on the JSON module if you want to 534Note that you are responsible to depend on the L<CBOR::XS> module if you
201use this functionality, as AnyEvent does not have a dependency itself. 535want to use this functionality, as AnyEvent does not have a dependency on
202 536it itself.
203=item filter_r => $cb
204
205=item filter_w => $cb
206
207These exist, but are undocumented at this time.
208 537
209=back 538=back
210 539
211=cut 540=cut
212 541
213sub new { 542sub new {
214 my $class = shift; 543 my $class = shift;
215
216 my $self = bless { @_ }, $class; 544 my $self = bless { @_ }, $class;
217 545
218 $self->{fh} or Carp::croak "mandatory argument fh is missing";
219
220 AnyEvent::Util::fh_nonblocking $self->{fh}, 1;
221
222 if ($self->{tls}) { 546 if ($self->{fh}) {
223 require Net::SSLeay;
224 $self->starttls (delete $self->{tls}, delete $self->{tls_ctx});
225 }
226
227# $self->on_eof (delete $self->{on_eof} ) if $self->{on_eof}; # nop
228# $self->on_error (delete $self->{on_error}) if $self->{on_error}; # nop
229# $self->on_read (delete $self->{on_read} ) if $self->{on_read}; # nop
230 $self->on_drain (delete $self->{on_drain}) if $self->{on_drain};
231
232 $self->{_activity} = AnyEvent->now;
233 $self->_timeout;
234
235 $self->start_read; 547 $self->_start;
548 return unless $self->{fh}; # could be gone by now
549
550 } elsif ($self->{connect}) {
551 require AnyEvent::Socket;
552
553 $self->{peername} = $self->{connect}[0]
554 unless exists $self->{peername};
555
556 $self->{_skip_drain_rbuf} = 1;
557
558 {
559 Scalar::Util::weaken (my $self = $self);
560
561 $self->{_connect} =
562 AnyEvent::Socket::tcp_connect (
563 $self->{connect}[0],
564 $self->{connect}[1],
565 sub {
566 my ($fh, $host, $port, $retry) = @_;
567
568 delete $self->{_connect}; # no longer needed
569
570 if ($fh) {
571 $self->{fh} = $fh;
572
573 delete $self->{_skip_drain_rbuf};
574 $self->_start;
575
576 $self->{on_connect}
577 and $self->{on_connect}($self, $host, $port, sub {
578 delete @$self{qw(fh _tw _rtw _wtw _ww _rw _eof _queue rbuf _wbuf tls _tls_rbuf _tls_wbuf)};
579 $self->{_skip_drain_rbuf} = 1;
580 &$retry;
581 });
582
583 } else {
584 if ($self->{on_connect_error}) {
585 $self->{on_connect_error}($self, "$!");
586 $self->destroy if $self;
587 } else {
588 $self->error ($!, 1);
589 }
590 }
591 },
592 sub {
593 local $self->{fh} = $_[0];
594
595 $self->{on_prepare}
596 ? $self->{on_prepare}->($self)
597 : ()
598 }
599 );
600 }
601
602 } else {
603 Carp::croak "AnyEvent::Handle: either an existing fh or the connect parameter must be specified";
604 }
236 605
237 $self 606 $self
238} 607}
239 608
240sub _shutdown { 609sub _start {
241 my ($self) = @_; 610 my ($self) = @_;
242 611
243 delete $self->{_tw}; 612 # too many clueless people try to use udp and similar sockets
244 delete $self->{_rw}; 613 # with AnyEvent::Handle, do them a favour.
245 delete $self->{_ww}; 614 my $type = getsockopt $self->{fh}, Socket::SOL_SOCKET (), Socket::SO_TYPE ();
246 delete $self->{fh}; 615 Carp::croak "AnyEvent::Handle: only stream sockets supported, anything else will NOT work!"
616 if Socket::SOCK_STREAM () != (unpack "I", $type) && defined $type;
617
618 AnyEvent::fh_unblock $self->{fh};
619
620 $self->{_activity} =
621 $self->{_ractivity} =
622 $self->{_wactivity} = AE::now;
623
624 $self->{read_size} ||= 2048;
625 $self->{max_read_size} = $self->{read_size}
626 if $self->{read_size} > ($self->{max_read_size} || MAX_READ_SIZE);
627
628 $self->timeout (delete $self->{timeout} ) if $self->{timeout};
629 $self->rtimeout (delete $self->{rtimeout} ) if $self->{rtimeout};
630 $self->wtimeout (delete $self->{wtimeout} ) if $self->{wtimeout};
631
632 $self->no_delay (delete $self->{no_delay} ) if exists $self->{no_delay} && $self->{no_delay};
633 $self->keepalive (delete $self->{keepalive}) if exists $self->{keepalive} && $self->{keepalive};
634
635 $self->oobinline (exists $self->{oobinline} ? delete $self->{oobinline} : 1);
636
637 $self->starttls (delete $self->{tls}, delete $self->{tls_ctx})
638 if $self->{tls};
639
640 $self->on_drain (delete $self->{on_drain} ) if $self->{on_drain};
641
642 $self->start_read
643 if $self->{on_read} || @{ $self->{_queue} };
644
645 $self->_drain_wbuf;
247} 646}
647
648=item $handle->error ($errno[, $fatal[, $message]])
649
650Generates an error event, just like AnyEvent::Handle itself would do, i.e.
651calls the C<on_error> callback.
652
653The only rerquired parameter is C<$errno>, which sets C<$!>. C<$fatal>
654defaults to false and C<$message> defaults to the stringified version
655of C<$1>.
656
657Example: generate C<EIO> when you read unexpected data.
658
659 $handle->push_read (line => sub {
660 $_[1] eq "hello"
661 or return $handle->error (Errno::EIO);
662 });
663
664=cut
248 665
249sub error { 666sub error {
250 my ($self) = @_; 667 my ($self, $errno, $fatal, $message) = @_;
251 668
252 { 669 $! = $errno;
253 local $!; 670 $message ||= "$!";
254 $self->_shutdown;
255 }
256 671
257 $self->{on_error}($self)
258 if $self->{on_error}; 672 if ($self->{on_error}) {
259 673 $self->{on_error}($self, $fatal, $message);
674 $self->destroy if $fatal;
675 } elsif ($self->{fh} || $self->{connect}) {
676 $self->destroy;
260 Carp::croak "AnyEvent::Handle uncaught fatal error: $!"; 677 Carp::croak "AnyEvent::Handle uncaught error: $message";
678 }
261} 679}
262 680
263=item $fh = $handle->fh 681=item $fh = $handle->fh
264 682
265This method returns the file handle of the L<AnyEvent::Handle> object. 683This method returns the file handle used to create the L<AnyEvent::Handle> object.
266 684
267=cut 685=cut
268 686
269sub fh { $_[0]{fh} } 687sub fh { $_[0]{fh} }
270 688
288 $_[0]{on_eof} = $_[1]; 706 $_[0]{on_eof} = $_[1];
289} 707}
290 708
291=item $handle->on_timeout ($cb) 709=item $handle->on_timeout ($cb)
292 710
293Replace the current C<on_timeout> callback, or disables the callback 711=item $handle->on_rtimeout ($cb)
294(but not the timeout) if C<$cb> = C<undef>. See C<timeout> constructor
295argument.
296 712
297=cut 713=item $handle->on_wtimeout ($cb)
298 714
715Replace the current C<on_timeout>, C<on_rtimeout> or C<on_wtimeout>
716callback, or disables the callback (but not the timeout) if C<$cb> =
717C<undef>. See the C<timeout> constructor argument and method.
718
719=cut
720
721# see below
722
723=item $handle->autocork ($boolean)
724
725Enables or disables the current autocork behaviour (see C<autocork>
726constructor argument). Changes will only take effect on the next write.
727
728=cut
729
730sub autocork {
731 $_[0]{autocork} = $_[1];
732}
733
734=item $handle->no_delay ($boolean)
735
736Enables or disables the C<no_delay> setting (see constructor argument of
737the same name for details).
738
739=cut
740
741sub no_delay {
742 $_[0]{no_delay} = $_[1];
743
744 setsockopt $_[0]{fh}, Socket::IPPROTO_TCP (), Socket::TCP_NODELAY (), int $_[1]
745 if $_[0]{fh};
746}
747
748=item $handle->keepalive ($boolean)
749
750Enables or disables the C<keepalive> setting (see constructor argument of
751the same name for details).
752
753=cut
754
755sub keepalive {
756 $_[0]{keepalive} = $_[1];
757
758 eval {
759 local $SIG{__DIE__};
760 setsockopt $_[0]{fh}, Socket::SOL_SOCKET (), Socket::SO_KEEPALIVE (), int $_[1]
761 if $_[0]{fh};
762 };
763}
764
765=item $handle->oobinline ($boolean)
766
767Enables or disables the C<oobinline> setting (see constructor argument of
768the same name for details).
769
770=cut
771
772sub oobinline {
773 $_[0]{oobinline} = $_[1];
774
775 eval {
776 local $SIG{__DIE__};
777 setsockopt $_[0]{fh}, Socket::SOL_SOCKET (), Socket::SO_OOBINLINE (), int $_[1]
778 if $_[0]{fh};
779 };
780}
781
782=item $handle->on_starttls ($cb)
783
784Replace the current C<on_starttls> callback (see the C<on_starttls> constructor argument).
785
786=cut
787
788sub on_starttls {
789 $_[0]{on_starttls} = $_[1];
790}
791
792=item $handle->on_stoptls ($cb)
793
794Replace the current C<on_stoptls> callback (see the C<on_stoptls> constructor argument).
795
796=cut
797
299sub on_timeout { 798sub on_stoptls {
300 $_[0]{on_timeout} = $_[1]; 799 $_[0]{on_stoptls} = $_[1];
800}
801
802=item $handle->rbuf_max ($max_octets)
803
804Configures the C<rbuf_max> setting (C<undef> disables it).
805
806=item $handle->wbuf_max ($max_octets)
807
808Configures the C<wbuf_max> setting (C<undef> disables it).
809
810=cut
811
812sub rbuf_max {
813 $_[0]{rbuf_max} = $_[1];
814}
815
816sub wbuf_max {
817 $_[0]{wbuf_max} = $_[1];
301} 818}
302 819
303############################################################################# 820#############################################################################
304 821
305=item $handle->timeout ($seconds) 822=item $handle->timeout ($seconds)
306 823
824=item $handle->rtimeout ($seconds)
825
826=item $handle->wtimeout ($seconds)
827
307Configures (or disables) the inactivity timeout. 828Configures (or disables) the inactivity timeout.
308 829
309=cut 830The timeout will be checked instantly, so this method might destroy the
831handle before it returns.
310 832
311sub timeout { 833=item $handle->timeout_reset
834
835=item $handle->rtimeout_reset
836
837=item $handle->wtimeout_reset
838
839Reset the activity timeout, as if data was received or sent.
840
841These methods are cheap to call.
842
843=cut
844
845for my $dir ("", "r", "w") {
846 my $timeout = "${dir}timeout";
847 my $tw = "_${dir}tw";
848 my $on_timeout = "on_${dir}timeout";
849 my $activity = "_${dir}activity";
850 my $cb;
851
852 *$on_timeout = sub {
853 $_[0]{$on_timeout} = $_[1];
854 };
855
856 *$timeout = sub {
312 my ($self, $timeout) = @_; 857 my ($self, $new_value) = @_;
313 858
859 $new_value >= 0
860 or Carp::croak "AnyEvent::Handle->$timeout called with negative timeout ($new_value), caught";
861
314 $self->{timeout} = $timeout; 862 $self->{$timeout} = $new_value;
315 $self->_timeout; 863 delete $self->{$tw}; &$cb;
316} 864 };
317 865
866 *{"${dir}timeout_reset"} = sub {
867 $_[0]{$activity} = AE::now;
868 };
869
870 # main workhorse:
318# reset the timeout watcher, as neccessary 871 # reset the timeout watcher, as neccessary
319# also check for time-outs 872 # also check for time-outs
320sub _timeout { 873 $cb = sub {
321 my ($self) = @_; 874 my ($self) = @_;
322 875
323 if ($self->{timeout}) { 876 if ($self->{$timeout} && $self->{fh}) {
324 my $NOW = AnyEvent->now; 877 my $NOW = AE::now;
325 878
326 # when would the timeout trigger? 879 # when would the timeout trigger?
327 my $after = $self->{_activity} + $self->{timeout} - $NOW; 880 my $after = $self->{$activity} + $self->{$timeout} - $NOW;
328 881
329 # now or in the past already? 882 # now or in the past already?
330 if ($after <= 0) { 883 if ($after <= 0) {
331 $self->{_activity} = $NOW; 884 $self->{$activity} = $NOW;
332 885
333 if ($self->{on_timeout}) { 886 if ($self->{$on_timeout}) {
334 $self->{on_timeout}($self); 887 $self->{$on_timeout}($self);
335 } else { 888 } else {
336 $! = Errno::ETIMEDOUT; 889 $self->error (Errno::ETIMEDOUT);
337 $self->error; 890 }
891
892 # callback could have changed timeout value, optimise
893 return unless $self->{$timeout};
894
895 # calculate new after
896 $after = $self->{$timeout};
338 } 897 }
339 898
340 # callbakx could have changed timeout value, optimise 899 Scalar::Util::weaken $self;
341 return unless $self->{timeout}; 900 return unless $self; # ->error could have destroyed $self
342 901
343 # calculate new after 902 $self->{$tw} ||= AE::timer $after, 0, sub {
344 $after = $self->{timeout}; 903 delete $self->{$tw};
904 $cb->($self);
905 };
906 } else {
907 delete $self->{$tw};
345 } 908 }
346
347 Scalar::Util::weaken $self;
348
349 $self->{_tw} ||= AnyEvent->timer (after => $after, cb => sub {
350 delete $self->{_tw};
351 $self->_timeout;
352 });
353 } else {
354 delete $self->{_tw};
355 } 909 }
356} 910}
357 911
358############################################################################# 912#############################################################################
359 913
366 920
367The write queue is very simple: you can add data to its end, and 921The write queue is very simple: you can add data to its end, and
368AnyEvent::Handle will automatically try to get rid of it for you. 922AnyEvent::Handle will automatically try to get rid of it for you.
369 923
370When data could be written and the write buffer is shorter then the low 924When data could be written and the write buffer is shorter then the low
371water mark, the C<on_drain> callback will be invoked. 925water mark, the C<on_drain> callback will be invoked once.
372 926
373=over 4 927=over 4
374 928
375=item $handle->on_drain ($cb) 929=item $handle->on_drain ($cb)
376 930
377Sets the C<on_drain> callback or clears it (see the description of 931Sets the C<on_drain> callback or clears it (see the description of
378C<on_drain> in the constructor). 932C<on_drain> in the constructor).
379 933
934This method may invoke callbacks (and therefore the handle might be
935destroyed after it returns).
936
380=cut 937=cut
381 938
382sub on_drain { 939sub on_drain {
383 my ($self, $cb) = @_; 940 my ($self, $cb) = @_;
384 941
385 $self->{on_drain} = $cb; 942 $self->{on_drain} = $cb;
386 943
387 $cb->($self) 944 $cb->($self)
388 if $cb && $self->{low_water_mark} >= length $self->{wbuf}; 945 if $cb && $self->{low_water_mark} >= (length $self->{wbuf}) + (length $self->{_tls_wbuf});
389} 946}
390 947
391=item $handle->push_write ($data) 948=item $handle->push_write ($data)
392 949
393Queues the given scalar to be written. You can push as much data as you 950Queues the given scalar to be written. You can push as much data as
394want (only limited by the available memory), as C<AnyEvent::Handle> 951you want (only limited by the available memory and C<wbuf_max>), as
395buffers it independently of the kernel. 952C<AnyEvent::Handle> buffers it independently of the kernel.
953
954This method may invoke callbacks (and therefore the handle might be
955destroyed after it returns).
396 956
397=cut 957=cut
398 958
399sub _drain_wbuf { 959sub _drain_wbuf {
400 my ($self) = @_; 960 my ($self) = @_;
404 Scalar::Util::weaken $self; 964 Scalar::Util::weaken $self;
405 965
406 my $cb = sub { 966 my $cb = sub {
407 my $len = syswrite $self->{fh}, $self->{wbuf}; 967 my $len = syswrite $self->{fh}, $self->{wbuf};
408 968
409 if ($len >= 0) { 969 if (defined $len) {
410 substr $self->{wbuf}, 0, $len, ""; 970 substr $self->{wbuf}, 0, $len, "";
411 971
412 $self->{_activity} = AnyEvent->now; 972 $self->{_activity} = $self->{_wactivity} = AE::now;
413 973
414 $self->{on_drain}($self) 974 $self->{on_drain}($self)
415 if $self->{low_water_mark} >= length $self->{wbuf} 975 if $self->{low_water_mark} >= (length $self->{wbuf}) + (length $self->{_tls_wbuf})
416 && $self->{on_drain}; 976 && $self->{on_drain};
417 977
418 delete $self->{_ww} unless length $self->{wbuf}; 978 delete $self->{_ww} unless length $self->{wbuf};
419 } elsif ($! != EAGAIN && $! != EINTR && $! != WSAEWOULDBLOCK) { 979 } elsif ($! != EAGAIN && $! != EINTR && $! != EWOULDBLOCK && $! != WSAEWOULDBLOCK) {
420 $self->error; 980 $self->error ($!, 1);
421 } 981 }
422 }; 982 };
423 983
424 # try to write data immediately 984 # try to write data immediately
425 $cb->(); 985 $cb->() unless $self->{autocork};
426 986
427 # if still data left in wbuf, we need to poll 987 # if still data left in wbuf, we need to poll
428 $self->{_ww} = AnyEvent->io (fh => $self->{fh}, poll => "w", cb => $cb) 988 $self->{_ww} = AE::io $self->{fh}, 1, $cb
429 if length $self->{wbuf}; 989 if length $self->{wbuf};
990
991 if (
992 defined $self->{wbuf_max}
993 && $self->{wbuf_max} < length $self->{wbuf}
994 ) {
995 $self->error (Errno::ENOSPC, 1), return;
996 }
430 }; 997 };
431} 998}
432 999
433our %WH; 1000our %WH;
434 1001
1002# deprecated
435sub register_write_type($$) { 1003sub register_write_type($$) {
436 $WH{$_[0]} = $_[1]; 1004 $WH{$_[0]} = $_[1];
437} 1005}
438 1006
439sub push_write { 1007sub push_write {
440 my $self = shift; 1008 my $self = shift;
441 1009
442 if (@_ > 1) { 1010 if (@_ > 1) {
443 my $type = shift; 1011 my $type = shift;
444 1012
1013 @_ = ($WH{$type} ||= _load_func "$type\::anyevent_write_type"
445 @_ = ($WH{$type} or Carp::croak "unsupported type passed to AnyEvent::Handle::push_write") 1014 or Carp::croak "unsupported/unloadable type '$type' passed to AnyEvent::Handle::push_write")
446 ->($self, @_); 1015 ->($self, @_);
447 } 1016 }
448 1017
1018 # we downgrade here to avoid hard-to-track-down bugs,
1019 # and diagnose the problem earlier and better.
1020
449 if ($self->{filter_w}) { 1021 if ($self->{tls}) {
450 $self->{filter_w}($self, \$_[0]); 1022 utf8::downgrade $self->{_tls_wbuf} .= $_[0];
1023 &_dotls ($self) if $self->{fh};
451 } else { 1024 } else {
452 $self->{wbuf} .= $_[0]; 1025 utf8::downgrade $self->{wbuf} .= $_[0];
453 $self->_drain_wbuf; 1026 $self->_drain_wbuf if $self->{fh};
454 } 1027 }
455} 1028}
456 1029
457=item $handle->push_write (type => @args) 1030=item $handle->push_write (type => @args)
458 1031
459=item $handle->unshift_write (type => @args)
460
461Instead of formatting your data yourself, you can also let this module do 1032Instead of formatting your data yourself, you can also let this module
462the job by specifying a type and type-specific arguments. 1033do the job by specifying a type and type-specific arguments. You
1034can also specify the (fully qualified) name of a package, in which
1035case AnyEvent tries to load the package and then expects to find the
1036C<anyevent_write_type> function inside (see "custom write types", below).
463 1037
464Predefined types are (if you have ideas for additional types, feel free to 1038Predefined types are (if you have ideas for additional types, feel free to
465drop by and tell us): 1039drop by and tell us):
466 1040
467=over 4 1041=over 4
469=item netstring => $string 1043=item netstring => $string
470 1044
471Formats the given value as netstring 1045Formats the given value as netstring
472(http://cr.yp.to/proto/netstrings.txt, this is not a recommendation to use them). 1046(http://cr.yp.to/proto/netstrings.txt, this is not a recommendation to use them).
473 1047
474=back
475
476=cut 1048=cut
477 1049
478register_write_type netstring => sub { 1050register_write_type netstring => sub {
479 my ($self, $string) = @_; 1051 my ($self, $string) = @_;
480 1052
481 sprintf "%d:%s,", (length $string), $string 1053 (length $string) . ":$string,"
1054};
1055
1056=item packstring => $format, $data
1057
1058An octet string prefixed with an encoded length. The encoding C<$format>
1059uses the same format as a Perl C<pack> format, but must specify a single
1060integer only (only one of C<cCsSlLqQiInNvVjJw> is allowed, plus an
1061optional C<!>, C<< < >> or C<< > >> modifier).
1062
1063=cut
1064
1065register_write_type packstring => sub {
1066 my ($self, $format, $string) = @_;
1067
1068 pack "$format/a*", $string
482}; 1069};
483 1070
484=item json => $array_or_hashref 1071=item json => $array_or_hashref
485 1072
486Encodes the given hash or array reference into a JSON object. Unless you 1073Encodes the given hash or array reference into a JSON object. Unless you
487provide your own JSON object, this means it will be encoded to JSON text 1074provide your own JSON object, this means it will be encoded to JSON text
488in UTF-8. 1075in UTF-8.
489 1076
1077The default encoder might or might not handle every type of JSON value -
1078it might be limited to arrays and objects for security reasons. See the
1079C<json> constructor attribute for more details.
1080
490JSON objects (and arrays) are self-delimiting, so you can write JSON at 1081JSON objects (and arrays) are self-delimiting, so if you only use arrays
491one end of a handle and read them at the other end without using any 1082and hashes, you can write JSON at one end of a handle and read them at the
492additional framing. 1083other end without using any additional framing.
493 1084
494The generated JSON text is guaranteed not to contain any newlines: While 1085The JSON text generated by the default encoder is guaranteed not to
495this module doesn't need delimiters after or between JSON texts to be 1086contain any newlines: While this module doesn't need delimiters after or
496able to read them, many other languages depend on that. 1087between JSON texts to be able to read them, many other languages depend on
1088them.
497 1089
498A simple RPC protocol that interoperates easily with others is to send 1090A simple RPC protocol that interoperates easily with other languages is
499JSON arrays (or objects, although arrays are usually the better choice as 1091to send JSON arrays (or objects, although arrays are usually the better
500they mimic how function argument passing works) and a newline after each 1092choice as they mimic how function argument passing works) and a newline
501JSON text: 1093after each JSON text:
502 1094
503 $handle->push_write (json => ["method", "arg1", "arg2"]); # whatever 1095 $handle->push_write (json => ["method", "arg1", "arg2"]); # whatever
504 $handle->push_write ("\012"); 1096 $handle->push_write ("\012");
505 1097
506An AnyEvent::Handle receiver would simply use the C<json> read type and 1098An AnyEvent::Handle receiver would simply use the C<json> read type and
509 $handle->push_read (json => sub { my $array = $_[1]; ... }); 1101 $handle->push_read (json => sub { my $array = $_[1]; ... });
510 1102
511Other languages could read single lines terminated by a newline and pass 1103Other languages could read single lines terminated by a newline and pass
512this line into their JSON decoder of choice. 1104this line into their JSON decoder of choice.
513 1105
1106=item cbor => $perl_scalar
1107
1108Encodes the given scalar into a CBOR value. Unless you provide your own
1109L<CBOR::XS> object, this means it will be encoded to a CBOR string not
1110using any extensions, if possible.
1111
1112CBOR values are self-delimiting, so you can write CBOR at one end of
1113a handle and read them at the other end without using any additional
1114framing.
1115
1116A simple nd very very fast RPC protocol that interoperates with
1117other languages is to send CBOR and receive CBOR values (arrays are
1118recommended):
1119
1120 $handle->push_write (cbor => ["method", "arg1", "arg2"]); # whatever
1121
1122An AnyEvent::Handle receiver would simply use the C<cbor> read type:
1123
1124 $handle->push_read (cbor => sub { my $array = $_[1]; ... });
1125
514=cut 1126=cut
1127
1128sub json_coder() {
1129 eval { require JSON::XS; JSON::XS->new->utf8 }
1130 || do { require JSON::PP; JSON::PP->new->utf8 }
1131}
515 1132
516register_write_type json => sub { 1133register_write_type json => sub {
517 my ($self, $ref) = @_; 1134 my ($self, $ref) = @_;
518 1135
519 require JSON; 1136 ($self->{json} ||= json_coder)
520 1137 ->encode ($ref)
521 $self->{json} ? $self->{json}->encode ($ref)
522 : JSON::encode_json ($ref)
523}; 1138};
524 1139
525=item AnyEvent::Handle::register_write_type type => $coderef->($handle, @args) 1140sub cbor_coder() {
1141 require CBOR::XS;
1142 CBOR::XS->new
1143}
526 1144
527This function (not method) lets you add your own types to C<push_write>. 1145register_write_type cbor => sub {
1146 my ($self, $scalar) = @_;
1147
1148 ($self->{cbor} ||= cbor_coder)
1149 ->encode ($scalar)
1150};
1151
1152=item storable => $reference
1153
1154Freezes the given reference using L<Storable> and writes it to the
1155handle. Uses the C<nfreeze> format.
1156
1157=cut
1158
1159register_write_type storable => sub {
1160 my ($self, $ref) = @_;
1161
1162 require Storable unless $Storable::VERSION;
1163
1164 pack "w/a*", Storable::nfreeze ($ref)
1165};
1166
1167=back
1168
1169=item $handle->push_shutdown
1170
1171Sometimes you know you want to close the socket after writing your data
1172before it was actually written. One way to do that is to replace your
1173C<on_drain> handler by a callback that shuts down the socket (and set
1174C<low_water_mark> to C<0>). This method is a shorthand for just that, and
1175replaces the C<on_drain> callback with:
1176
1177 sub { shutdown $_[0]{fh}, 1 }
1178
1179This simply shuts down the write side and signals an EOF condition to the
1180the peer.
1181
1182You can rely on the normal read queue and C<on_eof> handling
1183afterwards. This is the cleanest way to close a connection.
1184
1185This method may invoke callbacks (and therefore the handle might be
1186destroyed after it returns).
1187
1188=cut
1189
1190sub push_shutdown {
1191 my ($self) = @_;
1192
1193 delete $self->{low_water_mark};
1194 $self->on_drain (sub { shutdown $_[0]{fh}, 1 });
1195}
1196
1197=item custom write types - Package::anyevent_write_type $handle, @args
1198
1199Instead of one of the predefined types, you can also specify the name of
1200a package. AnyEvent will try to load the package and then expects to find
1201a function named C<anyevent_write_type> inside. If it isn't found, it
1202progressively tries to load the parent package until it either finds the
1203function (good) or runs out of packages (bad).
1204
528Whenever the given C<type> is used, C<push_write> will invoke the code 1205Whenever the given C<type> is used, C<push_write> will the function with
529reference with the handle object and the remaining arguments. 1206the handle object and the remaining arguments.
530 1207
531The code reference is supposed to return a single octet string that will 1208The function is supposed to return a single octet string that will be
532be appended to the write buffer. 1209appended to the write buffer, so you can mentally treat this function as a
1210"arguments to on-the-wire-format" converter.
533 1211
534Note that this is a function, and all types registered this way will be 1212Example: implement a custom write type C<join> that joins the remaining
535global, so try to use unique names. 1213arguments using the first one.
1214
1215 $handle->push_write (My::Type => " ", 1,2,3);
1216
1217 # uses the following package, which can be defined in the "My::Type" or in
1218 # the "My" modules to be auto-loaded, or just about anywhere when the
1219 # My::Type::anyevent_write_type is defined before invoking it.
1220
1221 package My::Type;
1222
1223 sub anyevent_write_type {
1224 my ($handle, $delim, @args) = @_;
1225
1226 join $delim, @args
1227 }
536 1228
537=cut 1229=cut
538 1230
539############################################################################# 1231#############################################################################
540 1232
549ways, the "simple" way, using only C<on_read> and the "complex" way, using 1241ways, the "simple" way, using only C<on_read> and the "complex" way, using
550a queue. 1242a queue.
551 1243
552In the simple case, you just install an C<on_read> callback and whenever 1244In the simple case, you just install an C<on_read> callback and whenever
553new data arrives, it will be called. You can then remove some data (if 1245new data arrives, it will be called. You can then remove some data (if
554enough is there) from the read buffer (C<< $handle->rbuf >>) if you want 1246enough is there) from the read buffer (C<< $handle->rbuf >>). Or you can
555or not. 1247leave the data there if you want to accumulate more (e.g. when only a
1248partial message has been received so far), or change the read queue with
1249e.g. C<push_read>.
556 1250
557In the more complex case, you want to queue multiple callbacks. In this 1251In the more complex case, you want to queue multiple callbacks. In this
558case, AnyEvent::Handle will call the first queued callback each time new 1252case, AnyEvent::Handle will call the first queued callback each time new
559data arrives and removes it when it has done its job (see C<push_read>, 1253data arrives (also the first time it is queued) and remove it when it has
560below). 1254done its job (see C<push_read>, below).
561 1255
562This way you can, for example, push three line-reads, followed by reading 1256This way you can, for example, push three line-reads, followed by reading
563a chunk of data, and AnyEvent::Handle will execute them in order. 1257a chunk of data, and AnyEvent::Handle will execute them in order.
564 1258
565Example 1: EPP protocol parser. EPP sends 4 byte length info, followed by 1259Example 1: EPP protocol parser. EPP sends 4 byte length info, followed by
566the specified number of bytes which give an XML datagram. 1260the specified number of bytes which give an XML datagram.
567 1261
568 # in the default state, expect some header bytes 1262 # in the default state, expect some header bytes
569 $handle->on_read (sub { 1263 $handle->on_read (sub {
570 # some data is here, now queue the length-header-read (4 octets) 1264 # some data is here, now queue the length-header-read (4 octets)
571 shift->unshift_read_chunk (4, sub { 1265 shift->unshift_read (chunk => 4, sub {
572 # header arrived, decode 1266 # header arrived, decode
573 my $len = unpack "N", $_[1]; 1267 my $len = unpack "N", $_[1];
574 1268
575 # now read the payload 1269 # now read the payload
576 shift->unshift_read_chunk ($len, sub { 1270 shift->unshift_read (chunk => $len, sub {
577 my $xml = $_[1]; 1271 my $xml = $_[1];
578 # handle xml 1272 # handle xml
579 }); 1273 });
580 }); 1274 });
581 }); 1275 });
582 1276
583Example 2: Implement a client for a protocol that replies either with 1277Example 2: Implement a client for a protocol that replies either with "OK"
584"OK" and another line or "ERROR" for one request, and 64 bytes for the 1278and another line or "ERROR" for the first request that is sent, and 64
585second request. Due tot he availability of a full queue, we can just 1279bytes for the second request. Due to the availability of a queue, we can
586pipeline sending both requests and manipulate the queue as necessary in 1280just pipeline sending both requests and manipulate the queue as necessary
587the callbacks: 1281in the callbacks.
588 1282
589 # request one 1283When the first callback is called and sees an "OK" response, it will
1284C<unshift> another line-read. This line-read will be queued I<before> the
128564-byte chunk callback.
1286
1287 # request one, returns either "OK + extra line" or "ERROR"
590 $handle->push_write ("request 1\015\012"); 1288 $handle->push_write ("request 1\015\012");
591 1289
592 # we expect "ERROR" or "OK" as response, so push a line read 1290 # we expect "ERROR" or "OK" as response, so push a line read
593 $handle->push_read_line (sub { 1291 $handle->push_read (line => sub {
594 # if we got an "OK", we have to _prepend_ another line, 1292 # if we got an "OK", we have to _prepend_ another line,
595 # so it will be read before the second request reads its 64 bytes 1293 # so it will be read before the second request reads its 64 bytes
596 # which are already in the queue when this callback is called 1294 # which are already in the queue when this callback is called
597 # we don't do this in case we got an error 1295 # we don't do this in case we got an error
598 if ($_[1] eq "OK") { 1296 if ($_[1] eq "OK") {
599 $_[0]->unshift_read_line (sub { 1297 $_[0]->unshift_read (line => sub {
600 my $response = $_[1]; 1298 my $response = $_[1];
601 ... 1299 ...
602 }); 1300 });
603 } 1301 }
604 }); 1302 });
605 1303
606 # request two 1304 # request two, simply returns 64 octets
607 $handle->push_write ("request 2\015\012"); 1305 $handle->push_write ("request 2\015\012");
608 1306
609 # simply read 64 bytes, always 1307 # simply read 64 bytes, always
610 $handle->push_read_chunk (64, sub { 1308 $handle->push_read (chunk => 64, sub {
611 my $response = $_[1]; 1309 my $response = $_[1];
612 ... 1310 ...
613 }); 1311 });
614 1312
615=over 4 1313=over 4
616 1314
617=cut 1315=cut
618 1316
619sub _drain_rbuf { 1317sub _drain_rbuf {
620 my ($self) = @_; 1318 my ($self) = @_;
1319
1320 # avoid recursion
1321 return if $self->{_skip_drain_rbuf};
1322 local $self->{_skip_drain_rbuf} = 1;
1323
1324 while () {
1325 # we need to use a separate tls read buffer, as we must not receive data while
1326 # we are draining the buffer, and this can only happen with TLS.
1327 $self->{rbuf} .= delete $self->{_tls_rbuf}
1328 if exists $self->{_tls_rbuf};
1329
1330 my $len = length $self->{rbuf};
1331
1332 if (my $cb = shift @{ $self->{_queue} }) {
1333 unless ($cb->($self)) {
1334 # no progress can be made
1335 # (not enough data and no data forthcoming)
1336 $self->error (Errno::EPIPE, 1), return
1337 if $self->{_eof};
1338
1339 unshift @{ $self->{_queue} }, $cb;
1340 last;
1341 }
1342 } elsif ($self->{on_read}) {
1343 last unless $len;
1344
1345 $self->{on_read}($self);
1346
1347 if (
1348 $len == length $self->{rbuf} # if no data has been consumed
1349 && !@{ $self->{_queue} } # and the queue is still empty
1350 && $self->{on_read} # but we still have on_read
1351 ) {
1352 # no further data will arrive
1353 # so no progress can be made
1354 $self->error (Errno::EPIPE, 1), return
1355 if $self->{_eof};
1356
1357 last; # more data might arrive
1358 }
1359 } else {
1360 # read side becomes idle
1361 delete $self->{_rw} unless $self->{tls};
1362 last;
1363 }
1364 }
1365
1366 if ($self->{_eof}) {
1367 $self->{on_eof}
1368 ? $self->{on_eof}($self)
1369 : $self->error (0, 1, "Unexpected end-of-file");
1370
1371 return;
1372 }
621 1373
622 if ( 1374 if (
623 defined $self->{rbuf_max} 1375 defined $self->{rbuf_max}
624 && $self->{rbuf_max} < length $self->{rbuf} 1376 && $self->{rbuf_max} < length $self->{rbuf}
625 ) { 1377 ) {
626 $! = &Errno::ENOSPC; 1378 $self->error (Errno::ENOSPC, 1), return;
627 $self->error;
628 } 1379 }
629 1380
630 return if $self->{in_drain}; 1381 # may need to restart read watcher
631 local $self->{in_drain} = 1; 1382 unless ($self->{_rw}) {
632 1383 $self->start_read
633 while (my $len = length $self->{rbuf}) { 1384 if $self->{on_read} || @{ $self->{_queue} };
634 no strict 'refs';
635 if (my $cb = shift @{ $self->{_queue} }) {
636 unless ($cb->($self)) {
637 if ($self->{_eof}) {
638 # no progress can be made (not enough data and no data forthcoming)
639 $! = &Errno::EPIPE;
640 $self->error;
641 }
642
643 unshift @{ $self->{_queue} }, $cb;
644 return;
645 }
646 } elsif ($self->{on_read}) {
647 $self->{on_read}($self);
648
649 if (
650 $self->{_eof} # if no further data will arrive
651 && $len == length $self->{rbuf} # and no data has been consumed
652 && !@{ $self->{_queue} } # and the queue is still empty
653 && $self->{on_read} # and we still want to read data
654 ) {
655 # then no progress can be made
656 $! = &Errno::EPIPE;
657 $self->error;
658 }
659 } else {
660 # read side becomes idle
661 delete $self->{_rw};
662 return;
663 }
664 } 1385 }
665
666 $self->{on_eof}($self)
667 if $self->{_eof} && $self->{on_eof};
668} 1386}
669 1387
670=item $handle->on_read ($cb) 1388=item $handle->on_read ($cb)
671 1389
672This replaces the currently set C<on_read> callback, or clears it (when 1390This replaces the currently set C<on_read> callback, or clears it (when
673the new callback is C<undef>). See the description of C<on_read> in the 1391the new callback is C<undef>). See the description of C<on_read> in the
674constructor. 1392constructor.
675 1393
1394This method may invoke callbacks (and therefore the handle might be
1395destroyed after it returns).
1396
676=cut 1397=cut
677 1398
678sub on_read { 1399sub on_read {
679 my ($self, $cb) = @_; 1400 my ($self, $cb) = @_;
680 1401
681 $self->{on_read} = $cb; 1402 $self->{on_read} = $cb;
1403 $self->_drain_rbuf if $cb;
682} 1404}
683 1405
684=item $handle->rbuf 1406=item $handle->rbuf
685 1407
686Returns the read buffer (as a modifiable lvalue). 1408Returns the read buffer (as a modifiable lvalue). You can also access the
1409read buffer directly as the C<< ->{rbuf} >> member, if you want (this is
1410much faster, and no less clean).
687 1411
688You can access the read buffer directly as the C<< ->{rbuf} >> member, if 1412The only operation allowed on the read buffer (apart from looking at it)
689you want. 1413is removing data from its beginning. Otherwise modifying or appending to
1414it is not allowed and will lead to hard-to-track-down bugs.
690 1415
691NOTE: The read buffer should only be used or modified if the C<on_read>, 1416NOTE: The read buffer should only be used or modified in the C<on_read>
692C<push_read> or C<unshift_read> methods are used. The other read methods 1417callback or when C<push_read> or C<unshift_read> are used with a single
693automatically manage the read buffer. 1418callback (i.e. untyped). Typed C<push_read> and C<unshift_read> methods
1419will manage the read buffer on their own.
694 1420
695=cut 1421=cut
696 1422
697sub rbuf : lvalue { 1423sub rbuf : lvalue {
698 $_[0]{rbuf} 1424 $_[0]{rbuf}
715 1441
716If enough data was available, then the callback must remove all data it is 1442If enough data was available, then the callback must remove all data it is
717interested in (which can be none at all) and return a true value. After returning 1443interested in (which can be none at all) and return a true value. After returning
718true, it will be removed from the queue. 1444true, it will be removed from the queue.
719 1445
1446These methods may invoke callbacks (and therefore the handle might be
1447destroyed after it returns).
1448
720=cut 1449=cut
721 1450
722our %RH; 1451our %RH;
723 1452
724sub register_read_type($$) { 1453sub register_read_type($$) {
730 my $cb = pop; 1459 my $cb = pop;
731 1460
732 if (@_) { 1461 if (@_) {
733 my $type = shift; 1462 my $type = shift;
734 1463
1464 $cb = ($RH{$type} ||= _load_func "$type\::anyevent_read_type"
735 $cb = ($RH{$type} or Carp::croak "unsupported type passed to AnyEvent::Handle::push_read") 1465 or Carp::croak "unsupported/unloadable type '$type' passed to AnyEvent::Handle::push_read")
736 ->($self, $cb, @_); 1466 ->($self, $cb, @_);
737 } 1467 }
738 1468
739 push @{ $self->{_queue} }, $cb; 1469 push @{ $self->{_queue} }, $cb;
740 $self->_drain_rbuf; 1470 $self->_drain_rbuf;
745 my $cb = pop; 1475 my $cb = pop;
746 1476
747 if (@_) { 1477 if (@_) {
748 my $type = shift; 1478 my $type = shift;
749 1479
1480 $cb = ($RH{$type} ||= _load_func "$type\::anyevent_read_type"
750 $cb = ($RH{$type} or Carp::croak "unsupported type passed to AnyEvent::Handle::unshift_read") 1481 or Carp::croak "unsupported/unloadable type '$type' passed to AnyEvent::Handle::unshift_read")
751 ->($self, $cb, @_); 1482 ->($self, $cb, @_);
752 } 1483 }
753
754 1484
755 unshift @{ $self->{_queue} }, $cb; 1485 unshift @{ $self->{_queue} }, $cb;
756 $self->_drain_rbuf; 1486 $self->_drain_rbuf;
757} 1487}
758 1488
760 1490
761=item $handle->unshift_read (type => @args, $cb) 1491=item $handle->unshift_read (type => @args, $cb)
762 1492
763Instead of providing a callback that parses the data itself you can chose 1493Instead of providing a callback that parses the data itself you can chose
764between a number of predefined parsing formats, for chunks of data, lines 1494between a number of predefined parsing formats, for chunks of data, lines
765etc. 1495etc. You can also specify the (fully qualified) name of a package, in
1496which case AnyEvent tries to load the package and then expects to find the
1497C<anyevent_read_type> function inside (see "custom read types", below).
766 1498
767Predefined types are (if you have ideas for additional types, feel free to 1499Predefined types are (if you have ideas for additional types, feel free to
768drop by and tell us): 1500drop by and tell us):
769 1501
770=over 4 1502=over 4
776data. 1508data.
777 1509
778Example: read 2 bytes. 1510Example: read 2 bytes.
779 1511
780 $handle->push_read (chunk => 2, sub { 1512 $handle->push_read (chunk => 2, sub {
781 warn "yay ", unpack "H*", $_[1]; 1513 say "yay " . unpack "H*", $_[1];
782 }); 1514 });
783 1515
784=cut 1516=cut
785 1517
786register_read_type chunk => sub { 1518register_read_type chunk => sub {
791 $cb->($_[0], substr $_[0]{rbuf}, 0, $len, ""); 1523 $cb->($_[0], substr $_[0]{rbuf}, 0, $len, "");
792 1 1524 1
793 } 1525 }
794}; 1526};
795 1527
796# compatibility with older API
797sub push_read_chunk {
798 $_[0]->push_read (chunk => $_[1], $_[2]);
799}
800
801sub unshift_read_chunk {
802 $_[0]->unshift_read (chunk => $_[1], $_[2]);
803}
804
805=item line => [$eol, ]$cb->($handle, $line, $eol) 1528=item line => [$eol, ]$cb->($handle, $line, $eol)
806 1529
807The callback will be called only once a full line (including the end of 1530The callback will be called only once a full line (including the end of
808line marker, C<$eol>) has been read. This line (excluding the end of line 1531line marker, C<$eol>) has been read. This line (excluding the end of line
809marker) will be passed to the callback as second argument (C<$line>), and 1532marker) will be passed to the callback as second argument (C<$line>), and
824=cut 1547=cut
825 1548
826register_read_type line => sub { 1549register_read_type line => sub {
827 my ($self, $cb, $eol) = @_; 1550 my ($self, $cb, $eol) = @_;
828 1551
829 $eol = qr|(\015?\012)| if @_ < 3; 1552 if (@_ < 3) {
1553 # this is faster then the generic code below
1554 sub {
1555 (my $pos = index $_[0]{rbuf}, "\012") >= 0
1556 or return;
1557
1558 (my $str = substr $_[0]{rbuf}, 0, $pos + 1, "") =~ s/(\015?\012)\Z// or die;
1559 $cb->($_[0], $str, "$1");
1560 1
1561 }
1562 } else {
830 $eol = quotemeta $eol unless ref $eol; 1563 $eol = quotemeta $eol unless ref $eol;
831 $eol = qr|^(.*?)($eol)|s; 1564 $eol = qr|^(.*?)($eol)|s;
832 1565
833 sub { 1566 sub {
834 $_[0]{rbuf} =~ s/$eol// or return; 1567 $_[0]{rbuf} =~ s/$eol// or return;
835 1568
836 $cb->($_[0], $1, $2); 1569 $cb->($_[0], "$1", "$2");
1570 1
837 1 1571 }
838 } 1572 }
839}; 1573};
840 1574
841# compatibility with older API
842sub push_read_line {
843 my $self = shift;
844 $self->push_read (line => @_);
845}
846
847sub unshift_read_line {
848 my $self = shift;
849 $self->unshift_read (line => @_);
850}
851
852=item netstring => $cb->($handle, $string)
853
854A netstring (http://cr.yp.to/proto/netstrings.txt, this is not an endorsement).
855
856Throws an error with C<$!> set to EBADMSG on format violations.
857
858=cut
859
860register_read_type netstring => sub {
861 my ($self, $cb) = @_;
862
863 sub {
864 unless ($_[0]{rbuf} =~ s/^(0|[1-9][0-9]*)://) {
865 if ($_[0]{rbuf} =~ /[^0-9]/) {
866 $! = &Errno::EBADMSG;
867 $self->error;
868 }
869 return;
870 }
871
872 my $len = $1;
873
874 $self->unshift_read (chunk => $len, sub {
875 my $string = $_[1];
876 $_[0]->unshift_read (chunk => 1, sub {
877 if ($_[1] eq ",") {
878 $cb->($_[0], $string);
879 } else {
880 $! = &Errno::EBADMSG;
881 $self->error;
882 }
883 });
884 });
885
886 1
887 }
888};
889
890=item regex => $accept[, $reject[, $skip], $cb->($handle, $data) 1575=item regex => $accept[, $reject[, $skip], $cb->($handle, $data)
891 1576
892Makes a regex match against the regex object C<$accept> and returns 1577Makes a regex match against the regex object C<$accept> and returns
893everything up to and including the match. 1578everything up to and including the match. All the usual regex variables
1579($1, %+ etc.) from the regex match are available in the callback.
894 1580
895Example: read a single line terminated by '\n'. 1581Example: read a single line terminated by '\n'.
896 1582
897 $handle->push_read (regex => qr<\n>, sub { ... }); 1583 $handle->push_read (regex => qr<\n>, sub { ... });
898 1584
911the receive buffer when neither C<$accept> nor C<$reject> match, 1597the receive buffer when neither C<$accept> nor C<$reject> match,
912and everything preceding and including the match will be accepted 1598and everything preceding and including the match will be accepted
913unconditionally. This is useful to skip large amounts of data that you 1599unconditionally. This is useful to skip large amounts of data that you
914know cannot be matched, so that the C<$accept> or C<$reject> regex do not 1600know cannot be matched, so that the C<$accept> or C<$reject> regex do not
915have to start matching from the beginning. This is purely an optimisation 1601have to start matching from the beginning. This is purely an optimisation
916and is usually worth only when you expect more than a few kilobytes. 1602and is usually worth it only when you expect more than a few kilobytes.
917 1603
918Example: expect a http header, which ends at C<\015\012\015\012>. Since we 1604Example: expect a http header, which ends at C<\015\012\015\012>. Since we
919expect the header to be very large (it isn't in practise, but...), we use 1605expect the header to be very large (it isn't in practice, but...), we use
920a skip regex to skip initial portions. The skip regex is tricky in that 1606a skip regex to skip initial portions. The skip regex is tricky in that
921it only accepts something not ending in either \015 or \012, as these are 1607it only accepts something not ending in either \015 or \012, as these are
922required for the accept regex. 1608required for the accept regex.
923 1609
924 $handle->push_read (regex => 1610 $handle->push_read (regex =>
937 1623
938 sub { 1624 sub {
939 # accept 1625 # accept
940 if ($$rbuf =~ $accept) { 1626 if ($$rbuf =~ $accept) {
941 $data .= substr $$rbuf, 0, $+[0], ""; 1627 $data .= substr $$rbuf, 0, $+[0], "";
942 $cb->($self, $data); 1628 $cb->($_[0], $data);
943 return 1; 1629 return 1;
944 } 1630 }
945 1631
946 # reject 1632 # reject
947 if ($reject && $$rbuf =~ $reject) { 1633 if ($reject && $$rbuf =~ $reject) {
948 $! = &Errno::EBADMSG; 1634 $_[0]->error (Errno::EBADMSG);
949 $self->error;
950 } 1635 }
951 1636
952 # skip 1637 # skip
953 if ($skip && $$rbuf =~ $skip) { 1638 if ($skip && $$rbuf =~ $skip) {
954 $data .= substr $$rbuf, 0, $+[0], ""; 1639 $data .= substr $$rbuf, 0, $+[0], "";
956 1641
957 () 1642 ()
958 } 1643 }
959}; 1644};
960 1645
1646=item netstring => $cb->($handle, $string)
1647
1648A netstring (http://cr.yp.to/proto/netstrings.txt, this is not an endorsement).
1649
1650Throws an error with C<$!> set to EBADMSG on format violations.
1651
1652=cut
1653
1654register_read_type netstring => sub {
1655 my ($self, $cb) = @_;
1656
1657 sub {
1658 unless ($_[0]{rbuf} =~ s/^(0|[1-9][0-9]*)://) {
1659 if ($_[0]{rbuf} =~ /[^0-9]/) {
1660 $_[0]->error (Errno::EBADMSG);
1661 }
1662 return;
1663 }
1664
1665 my $len = $1;
1666
1667 $_[0]->unshift_read (chunk => $len, sub {
1668 my $string = $_[1];
1669 $_[0]->unshift_read (chunk => 1, sub {
1670 if ($_[1] eq ",") {
1671 $cb->($_[0], $string);
1672 } else {
1673 $_[0]->error (Errno::EBADMSG);
1674 }
1675 });
1676 });
1677
1678 1
1679 }
1680};
1681
1682=item packstring => $format, $cb->($handle, $string)
1683
1684An octet string prefixed with an encoded length. The encoding C<$format>
1685uses the same format as a Perl C<pack> format, but must specify a single
1686integer only (only one of C<cCsSlLqQiInNvVjJw> is allowed, plus an
1687optional C<!>, C<< < >> or C<< > >> modifier).
1688
1689For example, DNS over TCP uses a prefix of C<n> (2 octet network order),
1690EPP uses a prefix of C<N> (4 octtes).
1691
1692Example: read a block of data prefixed by its length in BER-encoded
1693format (very efficient).
1694
1695 $handle->push_read (packstring => "w", sub {
1696 my ($handle, $data) = @_;
1697 });
1698
1699=cut
1700
1701register_read_type packstring => sub {
1702 my ($self, $cb, $format) = @_;
1703
1704 sub {
1705 # when we can use 5.10 we can use ".", but for 5.8 we use the re-pack method
1706 defined (my $len = eval { unpack $format, $_[0]{rbuf} })
1707 or return;
1708
1709 $format = length pack $format, $len;
1710
1711 # bypass unshift if we already have the remaining chunk
1712 if ($format + $len <= length $_[0]{rbuf}) {
1713 my $data = substr $_[0]{rbuf}, $format, $len;
1714 substr $_[0]{rbuf}, 0, $format + $len, "";
1715 $cb->($_[0], $data);
1716 } else {
1717 # remove prefix
1718 substr $_[0]{rbuf}, 0, $format, "";
1719
1720 # read remaining chunk
1721 $_[0]->unshift_read (chunk => $len, $cb);
1722 }
1723
1724 1
1725 }
1726};
1727
961=item json => $cb->($handle, $hash_or_arrayref) 1728=item json => $cb->($handle, $hash_or_arrayref)
962 1729
963Reads a JSON object or array, decodes it and passes it to the callback. 1730Reads a JSON object or array, decodes it and passes it to the
1731callback. When a parse error occurs, an C<EBADMSG> error will be raised.
964 1732
965If a C<json> object was passed to the constructor, then that will be used 1733If a C<json> object was passed to the constructor, then that will be
966for the final decode, otherwise it will create a JSON coder expecting UTF-8. 1734used for the final decode, otherwise it will create a L<JSON::XS> or
1735L<JSON::PP> coder object expecting UTF-8.
967 1736
968This read type uses the incremental parser available with JSON version 1737This read type uses the incremental parser available with JSON version
9692.09 (and JSON::XS version 2.2) and above. You have to provide a 17382.09 (and JSON::XS version 2.2) and above.
970dependency on your own: this module will load the JSON module, but
971AnyEvent does not depend on it itself.
972 1739
973Since JSON texts are fully self-delimiting, the C<json> read and write 1740Since JSON texts are fully self-delimiting, the C<json> read and write
974types are an ideal simple RPC protocol: just exchange JSON datagrams. See 1741types are an ideal simple RPC protocol: just exchange JSON datagrams. See
975the C<json> write type description, above, for an actual example. 1742the C<json> write type description, above, for an actual example.
976 1743
977=cut 1744=cut
978 1745
979register_read_type json => sub { 1746register_read_type json => sub {
980 my ($self, $cb, $accept, $reject, $skip) = @_; 1747 my ($self, $cb) = @_;
981 1748
982 require JSON; 1749 my $json = $self->{json} ||= json_coder;
983 1750
984 my $data; 1751 my $data;
985 my $rbuf = \$self->{rbuf};
986
987 my $json = $self->{json} ||= JSON->new->utf8;
988 1752
989 sub { 1753 sub {
990 my $ref = $json->incr_parse ($self->{rbuf}); 1754 my $ref = eval { $json->incr_parse ($_[0]{rbuf}) };
991 1755
992 if ($ref) { 1756 if ($ref) {
993 $self->{rbuf} = $json->incr_text; 1757 $_[0]{rbuf} = $json->incr_text;
994 $json->incr_text = ""; 1758 $json->incr_text = "";
995 $cb->($self, $ref); 1759 $cb->($_[0], $ref);
996 1760
997 1 1761 1
1762 } elsif ($@) {
1763 # error case
1764 $json->incr_skip;
1765
1766 $_[0]{rbuf} = $json->incr_text;
1767 $json->incr_text = "";
1768
1769 $_[0]->error (Errno::EBADMSG);
1770
1771 ()
998 } else { 1772 } else {
999 $self->{rbuf} = ""; 1773 $_[0]{rbuf} = "";
1774
1000 () 1775 ()
1001 } 1776 }
1002 } 1777 }
1003}; 1778};
1004 1779
1780=item cbor => $cb->($handle, $scalar)
1781
1782Reads a CBOR value, decodes it and passes it to the callback. When a parse
1783error occurs, an C<EBADMSG> error will be raised.
1784
1785If a L<CBOR::XS> object was passed to the constructor, then that will be
1786used for the final decode, otherwise it will create a CBOR coder without
1787enabling any options.
1788
1789You have to provide a dependency to L<CBOR::XS> on your own: this module
1790will load the L<CBOR::XS> module, but AnyEvent does not depend on it
1791itself.
1792
1793Since CBOR values are fully self-delimiting, the C<cbor> read and write
1794types are an ideal simple RPC protocol: just exchange CBOR datagrams. See
1795the C<cbor> write type description, above, for an actual example.
1796
1797=cut
1798
1799register_read_type cbor => sub {
1800 my ($self, $cb) = @_;
1801
1802 my $cbor = $self->{cbor} ||= cbor_coder;
1803
1804 my $data;
1805
1806 sub {
1807 my (@value) = eval { $cbor->incr_parse ($_[0]{rbuf}) };
1808
1809 if (@value) {
1810 $cb->($_[0], @value);
1811
1812 1
1813 } elsif ($@) {
1814 # error case
1815 $cbor->incr_reset;
1816
1817 $_[0]->error (Errno::EBADMSG);
1818
1819 ()
1820 } else {
1821 ()
1822 }
1823 }
1824};
1825
1826=item storable => $cb->($handle, $ref)
1827
1828Deserialises a L<Storable> frozen representation as written by the
1829C<storable> write type (BER-encoded length prefix followed by nfreeze'd
1830data).
1831
1832Raises C<EBADMSG> error if the data could not be decoded.
1833
1834=cut
1835
1836register_read_type storable => sub {
1837 my ($self, $cb) = @_;
1838
1839 require Storable unless $Storable::VERSION;
1840
1841 sub {
1842 # when we can use 5.10 we can use ".", but for 5.8 we use the re-pack method
1843 defined (my $len = eval { unpack "w", $_[0]{rbuf} })
1844 or return;
1845
1846 my $format = length pack "w", $len;
1847
1848 # bypass unshift if we already have the remaining chunk
1849 if ($format + $len <= length $_[0]{rbuf}) {
1850 my $data = substr $_[0]{rbuf}, $format, $len;
1851 substr $_[0]{rbuf}, 0, $format + $len, "";
1852
1853 eval { $cb->($_[0], Storable::thaw ($data)); 1 }
1854 or return $_[0]->error (Errno::EBADMSG);
1855 } else {
1856 # remove prefix
1857 substr $_[0]{rbuf}, 0, $format, "";
1858
1859 # read remaining chunk
1860 $_[0]->unshift_read (chunk => $len, sub {
1861 eval { $cb->($_[0], Storable::thaw ($_[1])); 1 }
1862 or $_[0]->error (Errno::EBADMSG);
1863 });
1864 }
1865
1866 1
1867 }
1868};
1869
1870=item tls_detect => $cb->($handle, $detect, $major, $minor)
1871
1872Checks the input stream for a valid SSL or TLS handshake TLSPaintext
1873record without consuming anything. Only SSL version 3 or higher
1874is handled, up to the fictituous protocol 4.x (but both SSL3+ and
1875SSL2-compatible framing is supported).
1876
1877If it detects that the input data is likely TLS, it calls the callback
1878with a true value for C<$detect> and the (on-wire) TLS version as second
1879and third argument (C<$major> is C<3>, and C<$minor> is 0..4 for SSL
18803.0, TLS 1.0, 1.1, 1.2 and 1.3, respectively). If it detects the input
1881to be definitely not TLS, it calls the callback with a false value for
1882C<$detect>.
1883
1884The callback could use this information to decide whether or not to start
1885TLS negotiation.
1886
1887In all cases the data read so far is passed to the following read
1888handlers.
1889
1890Usually you want to use the C<tls_autostart> read type instead.
1891
1892If you want to design a protocol that works in the presence of TLS
1893dtection, make sure that any non-TLS data doesn't start with the octet 22
1894(ASCII SYN, 16 hex) or 128-255 (i.e. highest bit set). The checks this
1895read type does are a bit more strict, but might losen in the future to
1896accomodate protocol changes.
1897
1898This read type does not rely on L<AnyEvent::TLS> (and thus, not on
1899L<Net::SSLeay>).
1900
1901=item tls_autostart => [$tls_ctx, ]$tls
1902
1903Tries to detect a valid SSL or TLS handshake. If one is detected, it tries
1904to start tls by calling C<starttls> with the given arguments.
1905
1906In practice, C<$tls> must be C<accept>, or a Net::SSLeay context that has
1907been configured to accept, as servers do not normally send a handshake on
1908their own and ths cannot be detected in this way.
1909
1910See C<tls_detect> above for more details.
1911
1912Example: give the client a chance to start TLS before accepting a text
1913line.
1914
1915 $hdl->push_read (tls_autostart => "accept");
1916 $hdl->push_read (line => sub {
1917 print "received ", ($_[0]{tls} ? "encrypted" : "cleartext"), " <$_[1]>\n";
1918 });
1919
1920=cut
1921
1922register_read_type tls_detect => sub {
1923 my ($self, $cb) = @_;
1924
1925 sub {
1926 # this regex matches a full or partial tls record
1927 if (
1928 # ssl3+: type(22=handshake) major(=3) minor(any) length_hi
1929 $self->{rbuf} =~ /^(?:\z| \x16 (\z| [\x03\x04] (?:\z| . (?:\z| [\x00-\x40] ))))/xs
1930 # ssl2 comapatible: len_hi len_lo type(1) major minor dummy(forlength)
1931 or $self->{rbuf} =~ /^(?:\z| [\x80-\xff] (?:\z| . (?:\z| \x01 (\z| [\x03\x04] (?:\z| . (?:\z| . ))))))/xs
1932 ) {
1933 return if 3 != length $1; # partial match, can't decide yet
1934
1935 # full match, valid TLS record
1936 my ($major, $minor) = unpack "CC", $1;
1937 $cb->($self, "accept", $major, $minor);
1938 } else {
1939 # mismatch == guaranteed not TLS
1940 $cb->($self, undef);
1941 }
1942
1943 1
1944 }
1945};
1946
1947register_read_type tls_autostart => sub {
1948 my ($self, @tls) = @_;
1949
1950 $RH{tls_detect}($self, sub {
1951 return unless $_[1];
1952 $_[0]->starttls (@tls);
1953 })
1954};
1955
1005=back 1956=back
1006 1957
1007=item AnyEvent::Handle::register_read_type type => $coderef->($handle, $cb, @args) 1958=item custom read types - Package::anyevent_read_type $handle, $cb, @args
1008 1959
1009This function (not method) lets you add your own types to C<push_read>. 1960Instead of one of the predefined types, you can also specify the name
1961of a package. AnyEvent will try to load the package and then expects to
1962find a function named C<anyevent_read_type> inside. If it isn't found, it
1963progressively tries to load the parent package until it either finds the
1964function (good) or runs out of packages (bad).
1010 1965
1011Whenever the given C<type> is used, C<push_read> will invoke the code 1966Whenever this type is used, C<push_read> will invoke the function with the
1012reference with the handle object, the callback and the remaining 1967handle object, the original callback and the remaining arguments.
1013arguments.
1014 1968
1015The code reference is supposed to return a callback (usually a closure) 1969The function is supposed to return a callback (usually a closure) that
1016that works as a plain read callback (see C<< ->push_read ($cb) >>). 1970works as a plain read callback (see C<< ->push_read ($cb) >>), so you can
1971mentally treat the function as a "configurable read type to read callback"
1972converter.
1017 1973
1018It should invoke the passed callback when it is done reading (remember to 1974It should invoke the original callback when it is done reading (remember
1019pass C<$handle> as first argument as all other callbacks do that). 1975to pass C<$handle> as first argument as all other callbacks do that,
1976although there is no strict requirement on this).
1020 1977
1021Note that this is a function, and all types registered this way will be
1022global, so try to use unique names.
1023
1024For examples, see the source of this module (F<perldoc -m AnyEvent::Handle>, 1978For examples, see the source of this module (F<perldoc -m
1025search for C<register_read_type>)). 1979AnyEvent::Handle>, search for C<register_read_type>)).
1026 1980
1027=item $handle->stop_read 1981=item $handle->stop_read
1028 1982
1029=item $handle->start_read 1983=item $handle->start_read
1030 1984
1031In rare cases you actually do not want to read anything from the 1985In rare cases you actually do not want to read anything from the
1032socket. In this case you can call C<stop_read>. Neither C<on_read> no 1986socket. In this case you can call C<stop_read>. Neither C<on_read> nor
1033any queued callbacks will be executed then. To start reading again, call 1987any queued callbacks will be executed then. To start reading again, call
1034C<start_read>. 1988C<start_read>.
1035 1989
1990Note that AnyEvent::Handle will automatically C<start_read> for you when
1991you change the C<on_read> callback or push/unshift a read callback, and it
1992will automatically C<stop_read> for you when neither C<on_read> is set nor
1993there are any read requests in the queue.
1994
1995In older versions of this module (<= 5.3), these methods had no effect,
1996as TLS does not support half-duplex connections. In current versions they
1997work as expected, as this behaviour is required to avoid certain resource
1998attacks, where the program would be forced to read (and buffer) arbitrary
1999amounts of data before being able to send some data. The drawback is that
2000some readings of the the SSL/TLS specifications basically require this
2001attack to be working, as SSL/TLS implementations might stall sending data
2002during a rehandshake.
2003
2004As a guideline, during the initial handshake, you should not stop reading,
2005and as a client, it might cause problems, depending on your application.
2006
1036=cut 2007=cut
1037 2008
1038sub stop_read { 2009sub stop_read {
1039 my ($self) = @_; 2010 my ($self) = @_;
1040 2011
1042} 2013}
1043 2014
1044sub start_read { 2015sub start_read {
1045 my ($self) = @_; 2016 my ($self) = @_;
1046 2017
1047 unless ($self->{_rw} || $self->{_eof}) { 2018 unless ($self->{_rw} || $self->{_eof} || !$self->{fh}) {
1048 Scalar::Util::weaken $self; 2019 Scalar::Util::weaken $self;
1049 2020
1050 $self->{_rw} = AnyEvent->io (fh => $self->{fh}, poll => "r", cb => sub { 2021 $self->{_rw} = AE::io $self->{fh}, 0, sub {
1051 my $rbuf = $self->{filter_r} ? \my $buf : \$self->{rbuf}; 2022 my $rbuf = \($self->{tls} ? my $buf : $self->{rbuf});
1052 my $len = sysread $self->{fh}, $$rbuf, $self->{read_size} || 8192, length $$rbuf; 2023 my $len = sysread $self->{fh}, $$rbuf, $self->{read_size}, length $$rbuf;
1053 2024
1054 if ($len > 0) { 2025 if ($len > 0) {
1055 $self->{_activity} = AnyEvent->now; 2026 $self->{_activity} = $self->{_ractivity} = AE::now;
1056 2027
1057 $self->{filter_r} 2028 if ($self->{tls}) {
1058 ? $self->{filter_r}($self, $rbuf) 2029 Net::SSLeay::BIO_write ($self->{_rbio}, $$rbuf);
2030
2031 &_dotls ($self);
2032 } else {
1059 : $self->_drain_rbuf; 2033 $self->_drain_rbuf;
2034 }
2035
2036 if ($len == $self->{read_size}) {
2037 $self->{read_size} *= 2;
2038 $self->{read_size} = $self->{max_read_size} || MAX_READ_SIZE
2039 if $self->{read_size} > ($self->{max_read_size} || MAX_READ_SIZE);
2040 }
1060 2041
1061 } elsif (defined $len) { 2042 } elsif (defined $len) {
1062 delete $self->{_rw}; 2043 delete $self->{_rw};
1063 $self->{_eof} = 1; 2044 $self->{_eof} = 1;
1064 $self->_drain_rbuf; 2045 $self->_drain_rbuf;
1065 2046
1066 } elsif ($! != EAGAIN && $! != EINTR && $! != WSAEWOULDBLOCK) { 2047 } elsif ($! != EAGAIN && $! != EINTR && $! != EWOULDBLOCK && $! != WSAEWOULDBLOCK) {
1067 return $self->error; 2048 return $self->error ($!, 1);
1068 } 2049 }
1069 }); 2050 };
1070 } 2051 }
1071} 2052}
1072 2053
2054our $ERROR_SYSCALL;
2055our $ERROR_WANT_READ;
2056
2057sub _tls_error {
2058 my ($self, $err) = @_;
2059
2060 return $self->error ($!, 1)
2061 if $err == Net::SSLeay::ERROR_SYSCALL ();
2062
2063 my $err = Net::SSLeay::ERR_error_string (Net::SSLeay::ERR_get_error ());
2064
2065 # reduce error string to look less scary
2066 $err =~ s/^error:[0-9a-fA-F]{8}:[^:]+:([^:]+):/\L$1: /;
2067
2068 if ($self->{_on_starttls}) {
2069 (delete $self->{_on_starttls})->($self, undef, $err);
2070 &_freetls;
2071 } else {
2072 &_freetls;
2073 $self->error (Errno::EPROTO, 1, $err);
2074 }
2075}
2076
2077# poll the write BIO and send the data if applicable
2078# also decode read data if possible
2079# this is basically our TLS state machine
2080# more efficient implementations are possible with openssl,
2081# but not with the buggy and incomplete Net::SSLeay.
1073sub _dotls { 2082sub _dotls {
1074 my ($self) = @_; 2083 my ($self) = @_;
1075 2084
2085 my $tmp;
2086
1076 if (length $self->{_tls_wbuf}) { 2087 while (length $self->{_tls_wbuf}) {
1077 while ((my $len = Net::SSLeay::write ($self->{tls}, $self->{_tls_wbuf})) > 0) { 2088 if (($tmp = Net::SSLeay::write ($self->{tls}, $self->{_tls_wbuf})) <= 0) {
1078 substr $self->{_tls_wbuf}, 0, $len, ""; 2089 $tmp = Net::SSLeay::get_error ($self->{tls}, $tmp);
2090
2091 return $self->_tls_error ($tmp)
2092 if $tmp != $ERROR_WANT_READ
2093 && ($tmp != $ERROR_SYSCALL || $!);
2094
2095 last;
1079 } 2096 }
1080 }
1081 2097
2098 substr $self->{_tls_wbuf}, 0, $tmp, "";
2099 }
2100
2101 while (defined ($tmp = Net::SSLeay::read ($self->{tls}))) {
2102 unless (length $tmp) {
2103 $self->{_on_starttls}
2104 and (delete $self->{_on_starttls})->($self, undef, "EOF during handshake"); # ???
2105 &_freetls;
2106
2107 if ($self->{on_stoptls}) {
2108 $self->{on_stoptls}($self);
2109 return;
2110 } else {
2111 # let's treat SSL-eof as we treat normal EOF
2112 delete $self->{_rw};
2113 $self->{_eof} = 1;
2114 }
2115 }
2116
2117 $self->{_tls_rbuf} .= $tmp;
2118 $self->_drain_rbuf;
2119 $self->{tls} or return; # tls session might have gone away in callback
2120 }
2121
2122 $tmp = Net::SSLeay::get_error ($self->{tls}, -1); # -1 is not neccessarily correct, but Net::SSLeay doesn't tell us
2123 return $self->_tls_error ($tmp)
2124 if $tmp != $ERROR_WANT_READ
2125 && ($tmp != $ERROR_SYSCALL || $!);
2126
1082 if (defined (my $buf = Net::SSLeay::BIO_read ($self->{_wbio}))) { 2127 while (length ($tmp = Net::SSLeay::BIO_read ($self->{_wbio}))) {
1083 $self->{wbuf} .= $buf; 2128 $self->{wbuf} .= $tmp;
1084 $self->_drain_wbuf; 2129 $self->_drain_wbuf;
2130 $self->{tls} or return; # tls session might have gone away in callback
1085 } 2131 }
1086 2132
1087 while (defined (my $buf = Net::SSLeay::read ($self->{tls}))) { 2133 $self->{_on_starttls}
1088 $self->{rbuf} .= $buf; 2134 and Net::SSLeay::state ($self->{tls}) == Net::SSLeay::ST_OK ()
1089 $self->_drain_rbuf; 2135 and (delete $self->{_on_starttls})->($self, 1, "TLS/SSL connection established");
1090 }
1091
1092 my $err = Net::SSLeay::get_error ($self->{tls}, -1);
1093
1094 if ($err!= Net::SSLeay::ERROR_WANT_READ ()) {
1095 if ($err == Net::SSLeay::ERROR_SYSCALL ()) {
1096 $self->error;
1097 } elsif ($err == Net::SSLeay::ERROR_SSL ()) {
1098 $! = &Errno::EIO;
1099 $self->error;
1100 }
1101
1102 # all others are fine for our purposes
1103 }
1104} 2136}
1105 2137
1106=item $handle->starttls ($tls[, $tls_ctx]) 2138=item $handle->starttls ($tls[, $tls_ctx])
1107 2139
1108Instead of starting TLS negotiation immediately when the AnyEvent::Handle 2140Instead of starting TLS negotiation immediately when the AnyEvent::Handle
1109object is created, you can also do that at a later time by calling 2141object is created, you can also do that at a later time by calling
1110C<starttls>. 2142C<starttls>. See the C<tls> constructor argument for general info.
2143
2144Starting TLS is currently an asynchronous operation - when you push some
2145write data and then call C<< ->starttls >> then TLS negotiation will start
2146immediately, after which the queued write data is then sent. This might
2147change in future versions, so best make sure you have no outstanding write
2148data when calling this method.
1111 2149
1112The first argument is the same as the C<tls> constructor argument (either 2150The first argument is the same as the C<tls> constructor argument (either
1113C<"connect">, C<"accept"> or an existing Net::SSLeay object). 2151C<"connect">, C<"accept"> or an existing Net::SSLeay object).
1114 2152
1115The second argument is the optional C<Net::SSLeay::CTX> object that is 2153The second argument is the optional C<AnyEvent::TLS> object that is used
1116used when AnyEvent::Handle has to create its own TLS connection object. 2154when AnyEvent::Handle has to create its own TLS connection object, or
2155a hash reference with C<< key => value >> pairs that will be used to
2156construct a new context.
1117 2157
1118The TLS connection object will end up in C<< $handle->{tls} >> after this 2158The TLS connection object will end up in C<< $handle->{tls} >>, the TLS
1119call and can be used or changed to your liking. Note that the handshake 2159context in C<< $handle->{tls_ctx} >> after this call and can be used or
1120might have already started when this function returns. 2160changed to your liking. Note that the handshake might have already started
2161when this function returns.
1121 2162
1122=cut 2163Due to bugs in OpenSSL, it might or might not be possible to do multiple
2164handshakes on the same stream. It is best to not attempt to use the
2165stream after stopping TLS.
1123 2166
1124# TODO: maybe document... 2167This method may invoke callbacks (and therefore the handle might be
2168destroyed after it returns).
2169
2170=cut
2171
2172our %TLS_CACHE; #TODO not yet documented, should we?
2173
1125sub starttls { 2174sub starttls {
1126 my ($self, $ssl, $ctx) = @_; 2175 my ($self, $tls, $ctx) = @_;
1127 2176
1128 $self->stoptls; 2177 Carp::croak "It is an error to call starttls on an AnyEvent::Handle object while TLS is already active, caught"
2178 if $self->{tls};
1129 2179
1130 if ($ssl eq "accept") { 2180 unless (defined $AnyEvent::TLS::VERSION) {
1131 $ssl = Net::SSLeay::new ($ctx || TLS_CTX ()); 2181 eval {
1132 Net::SSLeay::set_accept_state ($ssl); 2182 require Net::SSLeay;
1133 } elsif ($ssl eq "connect") { 2183 require AnyEvent::TLS;
1134 $ssl = Net::SSLeay::new ($ctx || TLS_CTX ()); 2184 1
1135 Net::SSLeay::set_connect_state ($ssl); 2185 } or return $self->error (Errno::EPROTO, 1, "TLS support not available on this system");
2186 }
2187
2188 $self->{tls} = $tls;
2189 $self->{tls_ctx} = $ctx if @_ > 2;
2190
2191 return unless $self->{fh};
2192
2193 $ERROR_SYSCALL = Net::SSLeay::ERROR_SYSCALL ();
2194 $ERROR_WANT_READ = Net::SSLeay::ERROR_WANT_READ ();
2195
2196 $tls = delete $self->{tls};
2197 $ctx = $self->{tls_ctx};
2198
2199 local $Carp::CarpLevel = 1; # skip ourselves when creating a new context or session
2200
2201 if ("HASH" eq ref $ctx) {
2202 if ($ctx->{cache}) {
2203 my $key = $ctx+0;
2204 $ctx = $TLS_CACHE{$key} ||= new AnyEvent::TLS %$ctx;
2205 } else {
2206 $ctx = new AnyEvent::TLS %$ctx;
2207 }
2208 }
1136 } 2209
1137 2210 $self->{tls_ctx} = $ctx || TLS_CTX ();
1138 $self->{tls} = $ssl; 2211 $self->{tls} = $tls = $self->{tls_ctx}->_get_session ($tls, $self, $self->{peername});
1139 2212
1140 # basically, this is deep magic (because SSL_read should have the same issues) 2213 # basically, this is deep magic (because SSL_read should have the same issues)
1141 # but the openssl maintainers basically said: "trust us, it just works". 2214 # but the openssl maintainers basically said: "trust us, it just works".
1142 # (unfortunately, we have to hardcode constants because the abysmally misdesigned 2215 # (unfortunately, we have to hardcode constants because the abysmally misdesigned
1143 # and mismaintained ssleay-module doesn't even offer them). 2216 # and mismaintained ssleay-module didn't offer them for a decade or so).
1144 # http://www.mail-archive.com/openssl-dev@openssl.org/msg22420.html 2217 # http://www.mail-archive.com/openssl-dev@openssl.org/msg22420.html
2218 #
2219 # in short: this is a mess.
2220 #
2221 # note that we do not try to keep the length constant between writes as we are required to do.
2222 # we assume that most (but not all) of this insanity only applies to non-blocking cases,
2223 # and we drive openssl fully in blocking mode here. Or maybe we don't - openssl seems to
2224 # have identity issues in that area.
1145 Net::SSLeay::CTX_set_mode ($self->{tls}, 2225# Net::SSLeay::set_mode ($ssl,
1146 (eval { local $SIG{__DIE__}; Net::SSLeay::MODE_ENABLE_PARTIAL_WRITE () } || 1) 2226# (eval { local $SIG{__DIE__}; Net::SSLeay::MODE_ENABLE_PARTIAL_WRITE () } || 1)
1147 | (eval { local $SIG{__DIE__}; Net::SSLeay::MODE_ACCEPT_MOVING_WRITE_BUFFER () } || 2)); 2227# | (eval { local $SIG{__DIE__}; Net::SSLeay::MODE_ACCEPT_MOVING_WRITE_BUFFER () } || 2));
2228 Net::SSLeay::set_mode ($tls, 1|2);
1148 2229
1149 $self->{_rbio} = Net::SSLeay::BIO_new (Net::SSLeay::BIO_s_mem ()); 2230 $self->{_rbio} = Net::SSLeay::BIO_new (Net::SSLeay::BIO_s_mem ());
1150 $self->{_wbio} = Net::SSLeay::BIO_new (Net::SSLeay::BIO_s_mem ()); 2231 $self->{_wbio} = Net::SSLeay::BIO_new (Net::SSLeay::BIO_s_mem ());
1151 2232
2233 Net::SSLeay::BIO_write ($self->{_rbio}, $self->{rbuf});
2234 $self->{rbuf} = "";
2235
1152 Net::SSLeay::set_bio ($ssl, $self->{_rbio}, $self->{_wbio}); 2236 Net::SSLeay::set_bio ($tls, $self->{_rbio}, $self->{_wbio});
1153 2237
1154 $self->{filter_w} = sub { 2238 $self->{_on_starttls} = sub { $_[0]{on_starttls}(@_) }
1155 $_[0]{_tls_wbuf} .= ${$_[1]}; 2239 if $self->{on_starttls};
1156 &_dotls; 2240
1157 }; 2241 &_dotls; # need to trigger the initial handshake
1158 $self->{filter_r} = sub { 2242 $self->start_read; # make sure we actually do read
1159 Net::SSLeay::BIO_write ($_[0]{_rbio}, ${$_[1]});
1160 &_dotls;
1161 };
1162} 2243}
1163 2244
1164=item $handle->stoptls 2245=item $handle->stoptls
1165 2246
1166Destroys the SSL connection, if any. Partial read or write data will be 2247Shuts down the SSL connection - this makes a proper EOF handshake by
1167lost. 2248sending a close notify to the other side, but since OpenSSL doesn't
2249support non-blocking shut downs, it is not guaranteed that you can re-use
2250the stream afterwards.
2251
2252This method may invoke callbacks (and therefore the handle might be
2253destroyed after it returns).
1168 2254
1169=cut 2255=cut
1170 2256
1171sub stoptls { 2257sub stoptls {
1172 my ($self) = @_; 2258 my ($self) = @_;
1173 2259
1174 Net::SSLeay::free (delete $self->{tls}) if $self->{tls}; 2260 if ($self->{tls} && $self->{fh}) {
2261 Net::SSLeay::shutdown ($self->{tls});
1175 2262
1176 delete $self->{_rbio}; 2263 &_dotls;
1177 delete $self->{_wbio}; 2264
1178 delete $self->{_tls_wbuf}; 2265# # we don't give a shit. no, we do, but we can't. no...#d#
1179 delete $self->{filter_r}; 2266# # we, we... have to use openssl :/#d#
1180 delete $self->{filter_w}; 2267# &_freetls;#d#
2268 }
1181} 2269}
2270
2271sub _freetls {
2272 my ($self) = @_;
2273
2274 return unless $self->{tls};
2275
2276 $self->{tls_ctx}->_put_session (delete $self->{tls})
2277 if $self->{tls} > 0;
2278
2279 delete @$self{qw(_rbio _wbio _tls_wbuf _on_starttls)};
2280}
2281
2282=item $handle->resettls
2283
2284This rarely-used method simply resets and TLS state on the handle, usually
2285causing data loss.
2286
2287One case where it may be useful is when you want to skip over the data in
2288the stream but you are not interested in interpreting it, so data loss is
2289no concern.
2290
2291=cut
2292
2293*resettls = \&_freetls;
1182 2294
1183sub DESTROY { 2295sub DESTROY {
1184 my $self = shift; 2296 my ($self) = @_;
1185 2297
1186 $self->stoptls; 2298 &_freetls;
2299
2300 my $linger = exists $self->{linger} ? $self->{linger} : 3600;
2301
2302 if ($linger && length $self->{wbuf} && $self->{fh}) {
2303 my $fh = delete $self->{fh};
2304 my $wbuf = delete $self->{wbuf};
2305
2306 my @linger;
2307
2308 push @linger, AE::io $fh, 1, sub {
2309 my $len = syswrite $fh, $wbuf, length $wbuf;
2310
2311 if ($len > 0) {
2312 substr $wbuf, 0, $len, "";
2313 } elsif (defined $len || ($! != EAGAIN && $! != EINTR && $! != EWOULDBLOCK && $! != WSAEWOULDBLOCK)) {
2314 @linger = (); # end
2315 }
2316 };
2317 push @linger, AE::timer $linger, 0, sub {
2318 @linger = ();
2319 };
2320 }
1187} 2321}
2322
2323=item $handle->destroy
2324
2325Shuts down the handle object as much as possible - this call ensures that
2326no further callbacks will be invoked and as many resources as possible
2327will be freed. Any method you will call on the handle object after
2328destroying it in this way will be silently ignored (and it will return the
2329empty list).
2330
2331Normally, you can just "forget" any references to an AnyEvent::Handle
2332object and it will simply shut down. This works in fatal error and EOF
2333callbacks, as well as code outside. It does I<NOT> work in a read or write
2334callback, so when you want to destroy the AnyEvent::Handle object from
2335within such an callback. You I<MUST> call C<< ->destroy >> explicitly in
2336that case.
2337
2338Destroying the handle object in this way has the advantage that callbacks
2339will be removed as well, so if those are the only reference holders (as
2340is common), then one doesn't need to do anything special to break any
2341reference cycles.
2342
2343The handle might still linger in the background and write out remaining
2344data, as specified by the C<linger> option, however.
2345
2346=cut
2347
2348sub destroy {
2349 my ($self) = @_;
2350
2351 $self->DESTROY;
2352 %$self = ();
2353 bless $self, "AnyEvent::Handle::destroyed";
2354}
2355
2356sub AnyEvent::Handle::destroyed::AUTOLOAD {
2357 #nop
2358}
2359
2360=item $handle->destroyed
2361
2362Returns false as long as the handle hasn't been destroyed by a call to C<<
2363->destroy >>, true otherwise.
2364
2365Can be useful to decide whether the handle is still valid after some
2366callback possibly destroyed the handle. For example, C<< ->push_write >>,
2367C<< ->starttls >> and other methods can call user callbacks, which in turn
2368can destroy the handle, so work can be avoided by checking sometimes:
2369
2370 $hdl->starttls ("accept");
2371 return if $hdl->destroyed;
2372 $hdl->push_write (...
2373
2374Note that the call to C<push_write> will silently be ignored if the handle
2375has been destroyed, so often you can just ignore the possibility of the
2376handle being destroyed.
2377
2378=cut
2379
2380sub destroyed { 0 }
2381sub AnyEvent::Handle::destroyed::destroyed { 1 }
1188 2382
1189=item AnyEvent::Handle::TLS_CTX 2383=item AnyEvent::Handle::TLS_CTX
1190 2384
1191This function creates and returns the Net::SSLeay::CTX object used by 2385This function creates and returns the AnyEvent::TLS object used by default
1192default for TLS mode. 2386for TLS mode.
1193 2387
1194The context is created like this: 2388The context is created by calling L<AnyEvent::TLS> without any arguments.
1195
1196 Net::SSLeay::load_error_strings;
1197 Net::SSLeay::SSLeay_add_ssl_algorithms;
1198 Net::SSLeay::randomize;
1199
1200 my $CTX = Net::SSLeay::CTX_new;
1201
1202 Net::SSLeay::CTX_set_options $CTX, Net::SSLeay::OP_ALL
1203 2389
1204=cut 2390=cut
1205 2391
1206our $TLS_CTX; 2392our $TLS_CTX;
1207 2393
1208sub TLS_CTX() { 2394sub TLS_CTX() {
1209 $TLS_CTX || do { 2395 $TLS_CTX ||= do {
1210 require Net::SSLeay; 2396 require AnyEvent::TLS;
1211 2397
1212 Net::SSLeay::load_error_strings (); 2398 new AnyEvent::TLS
1213 Net::SSLeay::SSLeay_add_ssl_algorithms (); 2399 }
1214 Net::SSLeay::randomize (); 2400}
1215 2401
1216 $TLS_CTX = Net::SSLeay::CTX_new (); 2402=back
1217 2403
1218 Net::SSLeay::CTX_set_options ($TLS_CTX, Net::SSLeay::OP_ALL ());
1219 2404
1220 $TLS_CTX 2405=head1 NONFREQUENTLY ASKED QUESTIONS
2406
2407=over 4
2408
2409=item I C<undef> the AnyEvent::Handle reference inside my callback and
2410still get further invocations!
2411
2412That's because AnyEvent::Handle keeps a reference to itself when handling
2413read or write callbacks.
2414
2415It is only safe to "forget" the reference inside EOF or error callbacks,
2416from within all other callbacks, you need to explicitly call the C<<
2417->destroy >> method.
2418
2419=item Why is my C<on_eof> callback never called?
2420
2421Probably because your C<on_error> callback is being called instead: When
2422you have outstanding requests in your read queue, then an EOF is
2423considered an error as you clearly expected some data.
2424
2425To avoid this, make sure you have an empty read queue whenever your handle
2426is supposed to be "idle" (i.e. connection closes are O.K.). You can set
2427an C<on_read> handler that simply pushes the first read requests in the
2428queue.
2429
2430See also the next question, which explains this in a bit more detail.
2431
2432=item How can I serve requests in a loop?
2433
2434Most protocols consist of some setup phase (authentication for example)
2435followed by a request handling phase, where the server waits for requests
2436and handles them, in a loop.
2437
2438There are two important variants: The first (traditional, better) variant
2439handles requests until the server gets some QUIT command, causing it to
2440close the connection first (highly desirable for a busy TCP server). A
2441client dropping the connection is an error, which means this variant can
2442detect an unexpected detection close.
2443
2444To handle this case, always make sure you have a non-empty read queue, by
2445pushing the "read request start" handler on it:
2446
2447 # we assume a request starts with a single line
2448 my @start_request; @start_request = (line => sub {
2449 my ($hdl, $line) = @_;
2450
2451 ... handle request
2452
2453 # push next request read, possibly from a nested callback
2454 $hdl->push_read (@start_request);
2455 });
2456
2457 # auth done, now go into request handling loop
2458 # now push the first @start_request
2459 $hdl->push_read (@start_request);
2460
2461By always having an outstanding C<push_read>, the handle always expects
2462some data and raises the C<EPIPE> error when the connction is dropped
2463unexpectedly.
2464
2465The second variant is a protocol where the client can drop the connection
2466at any time. For TCP, this means that the server machine may run out of
2467sockets easier, and in general, it means you cannot distinguish a protocl
2468failure/client crash from a normal connection close. Nevertheless, these
2469kinds of protocols are common (and sometimes even the best solution to the
2470problem).
2471
2472Having an outstanding read request at all times is possible if you ignore
2473C<EPIPE> errors, but this doesn't help with when the client drops the
2474connection during a request, which would still be an error.
2475
2476A better solution is to push the initial request read in an C<on_read>
2477callback. This avoids an error, as when the server doesn't expect data
2478(i.e. is idly waiting for the next request, an EOF will not raise an
2479error, but simply result in an C<on_eof> callback. It is also a bit slower
2480and simpler:
2481
2482 # auth done, now go into request handling loop
2483 $hdl->on_read (sub {
2484 my ($hdl) = @_;
2485
2486 # called each time we receive data but the read queue is empty
2487 # simply start read the request
2488
2489 $hdl->push_read (line => sub {
2490 my ($hdl, $line) = @_;
2491
2492 ... handle request
2493
2494 # do nothing special when the request has been handled, just
2495 # let the request queue go empty.
2496 });
2497 });
2498
2499=item I get different callback invocations in TLS mode/Why can't I pause
2500reading?
2501
2502Unlike, say, TCP, TLS connections do not consist of two independent
2503communication channels, one for each direction. Or put differently, the
2504read and write directions are not independent of each other: you cannot
2505write data unless you are also prepared to read, and vice versa.
2506
2507This means that, in TLS mode, you might get C<on_error> or C<on_eof>
2508callback invocations when you are not expecting any read data - the reason
2509is that AnyEvent::Handle always reads in TLS mode.
2510
2511During the connection, you have to make sure that you always have a
2512non-empty read-queue, or an C<on_read> watcher. At the end of the
2513connection (or when you no longer want to use it) you can call the
2514C<destroy> method.
2515
2516=item How do I read data until the other side closes the connection?
2517
2518If you just want to read your data into a perl scalar, the easiest way
2519to achieve this is by setting an C<on_read> callback that does nothing,
2520clearing the C<on_eof> callback and in the C<on_error> callback, the data
2521will be in C<$_[0]{rbuf}>:
2522
2523 $handle->on_read (sub { });
2524 $handle->on_eof (undef);
2525 $handle->on_error (sub {
2526 my $data = delete $_[0]{rbuf};
2527 });
2528
2529Note that this example removes the C<rbuf> member from the handle object,
2530which is not normally allowed by the API. It is expressly permitted in
2531this case only, as the handle object needs to be destroyed afterwards.
2532
2533The reason to use C<on_error> is that TCP connections, due to latencies
2534and packets loss, might get closed quite violently with an error, when in
2535fact all data has been received.
2536
2537It is usually better to use acknowledgements when transferring data,
2538to make sure the other side hasn't just died and you got the data
2539intact. This is also one reason why so many internet protocols have an
2540explicit QUIT command.
2541
2542=item I don't want to destroy the handle too early - how do I wait until
2543all data has been written?
2544
2545After writing your last bits of data, set the C<on_drain> callback
2546and destroy the handle in there - with the default setting of
2547C<low_water_mark> this will be called precisely when all data has been
2548written to the socket:
2549
2550 $handle->push_write (...);
2551 $handle->on_drain (sub {
2552 AE::log debug => "All data submitted to the kernel.";
2553 undef $handle;
2554 });
2555
2556If you just want to queue some data and then signal EOF to the other side,
2557consider using C<< ->push_shutdown >> instead.
2558
2559=item I want to contact a TLS/SSL server, I don't care about security.
2560
2561If your TLS server is a pure TLS server (e.g. HTTPS) that only speaks TLS,
2562connect to it and then create the AnyEvent::Handle with the C<tls>
2563parameter:
2564
2565 tcp_connect $host, $port, sub {
2566 my ($fh) = @_;
2567
2568 my $handle = new AnyEvent::Handle
2569 fh => $fh,
2570 tls => "connect",
2571 on_error => sub { ... };
2572
2573 $handle->push_write (...);
1221 } 2574 };
1222} 2575
2576=item I want to contact a TLS/SSL server, I do care about security.
2577
2578Then you should additionally enable certificate verification, including
2579peername verification, if the protocol you use supports it (see
2580L<AnyEvent::TLS>, C<verify_peername>).
2581
2582E.g. for HTTPS:
2583
2584 tcp_connect $host, $port, sub {
2585 my ($fh) = @_;
2586
2587 my $handle = new AnyEvent::Handle
2588 fh => $fh,
2589 peername => $host,
2590 tls => "connect",
2591 tls_ctx => { verify => 1, verify_peername => "https" },
2592 ...
2593
2594Note that you must specify the hostname you connected to (or whatever
2595"peername" the protocol needs) as the C<peername> argument, otherwise no
2596peername verification will be done.
2597
2598The above will use the system-dependent default set of trusted CA
2599certificates. If you want to check against a specific CA, add the
2600C<ca_file> (or C<ca_cert>) arguments to C<tls_ctx>:
2601
2602 tls_ctx => {
2603 verify => 1,
2604 verify_peername => "https",
2605 ca_file => "my-ca-cert.pem",
2606 },
2607
2608=item I want to create a TLS/SSL server, how do I do that?
2609
2610Well, you first need to get a server certificate and key. You have
2611three options: a) ask a CA (buy one, use cacert.org etc.) b) create a
2612self-signed certificate (cheap. check the search engine of your choice,
2613there are many tutorials on the net) or c) make your own CA (tinyca2 is a
2614nice program for that purpose).
2615
2616Then create a file with your private key (in PEM format, see
2617L<AnyEvent::TLS>), followed by the certificate (also in PEM format). The
2618file should then look like this:
2619
2620 -----BEGIN RSA PRIVATE KEY-----
2621 ...header data
2622 ... lots of base64'y-stuff
2623 -----END RSA PRIVATE KEY-----
2624
2625 -----BEGIN CERTIFICATE-----
2626 ... lots of base64'y-stuff
2627 -----END CERTIFICATE-----
2628
2629The important bits are the "PRIVATE KEY" and "CERTIFICATE" parts. Then
2630specify this file as C<cert_file>:
2631
2632 tcp_server undef, $port, sub {
2633 my ($fh) = @_;
2634
2635 my $handle = new AnyEvent::Handle
2636 fh => $fh,
2637 tls => "accept",
2638 tls_ctx => { cert_file => "my-server-keycert.pem" },
2639 ...
2640
2641When you have intermediate CA certificates that your clients might not
2642know about, just append them to the C<cert_file>.
1223 2643
1224=back 2644=back
1225 2645
1226=head1 SUBCLASSING AnyEvent::Handle 2646=head1 SUBCLASSING AnyEvent::Handle
1227 2647
1233=over 4 2653=over 4
1234 2654
1235=item * all constructor arguments become object members. 2655=item * all constructor arguments become object members.
1236 2656
1237At least initially, when you pass a C<tls>-argument to the constructor it 2657At least initially, when you pass a C<tls>-argument to the constructor it
1238will end up in C<< $handle->{tls} >>. Those members might be changes or 2658will end up in C<< $handle->{tls} >>. Those members might be changed or
1239mutated later on (for example C<tls> will hold the TLS connection object). 2659mutated later on (for example C<tls> will hold the TLS connection object).
1240 2660
1241=item * other object member names are prefixed with an C<_>. 2661=item * other object member names are prefixed with an C<_>.
1242 2662
1243All object members not explicitly documented (internal use) are prefixed 2663All object members not explicitly documented (internal use) are prefixed
1246 2666
1247=item * all members not documented here and not prefixed with an underscore 2667=item * all members not documented here and not prefixed with an underscore
1248are free to use in subclasses. 2668are free to use in subclasses.
1249 2669
1250Of course, new versions of AnyEvent::Handle may introduce more "public" 2670Of course, new versions of AnyEvent::Handle may introduce more "public"
1251member variables, but thats just life, at least it is documented. 2671member variables, but that's just life. At least it is documented.
1252 2672
1253=back 2673=back
1254 2674
1255=head1 AUTHOR 2675=head1 AUTHOR
1256 2676
1257Robin Redeker C<< <elmex at ta-sa.org> >>, Marc Lehmann <schmorp@schmorp.de>. 2677Robin Redeker C<< <elmex at ta-sa.org> >>, Marc Lehmann <schmorp@schmorp.de>.
1258 2678
1259=cut 2679=cut
1260 2680
12611; # End of AnyEvent::Handle 26811
2682

Diff Legend

Removed lines
+ Added lines
< Changed lines
> Changed lines