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 1.79 by root, Sun Jul 27 08:37:56 2008 UTC vs.
Revision 1.206 by root, Mon Nov 15 19:49:31 2010 UTC

1package AnyEvent::Handle;
2
3no warnings;
4use strict qw(subs vars);
5
6use AnyEvent ();
7use AnyEvent::Util qw(WSAEWOULDBLOCK);
8use Scalar::Util ();
9use Carp ();
10use Fcntl ();
11use Errno qw(EAGAIN EINTR);
12
13=head1 NAME 1=head1 NAME
14 2
15AnyEvent::Handle - non-blocking I/O on file handles via AnyEvent 3AnyEvent::Handle - non-blocking I/O on streaming handles via AnyEvent
16
17=cut
18
19our $VERSION = 4.22;
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 warn "got error $msg\n";
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 warn "got line <$line>\n";
43 $cv->send; 28 $cv->send;
44 }); 29 });
45 30
46 $cv->recv; 31 $cv->recv;
47 32
48=head1 DESCRIPTION 33=head1 DESCRIPTION
49 34
50This module is a helper module to make it easier to do event-based I/O on 35This is a helper module to make it easier to do event-based I/O on
51filehandles. For utility functions for doing non-blocking connects and accepts 36stream-based filehandles (sockets, pipes, and other stream things).
52on sockets see L<AnyEvent::Util>.
53 37
38The L<AnyEvent::Intro> tutorial contains some well-documented
39AnyEvent::Handle examples.
40
54In the following, when the documentation refers to of "bytes" then this 41In the following, where the documentation refers to "bytes", it means
55means characters. As sysread and syswrite are used for all I/O, their 42characters. As sysread and syswrite are used for all I/O, their
56treatment of characters applies to this module as well. 43treatment of characters applies to this module as well.
44
45At the very minimum, you should specify C<fh> or C<connect>, and the
46C<on_error> callback.
57 47
58All callbacks will be invoked with the handle object as their first 48All callbacks will be invoked with the handle object as their first
59argument. 49argument.
60 50
51=cut
52
53package AnyEvent::Handle;
54
55use Scalar::Util ();
56use List::Util ();
57use Carp ();
58use Errno qw(EAGAIN EINTR);
59
60use AnyEvent (); BEGIN { AnyEvent::common_sense }
61use AnyEvent::Util qw(WSAEWOULDBLOCK);
62
63our $VERSION = $AnyEvent::VERSION;
64
65sub _load_func($) {
66 my $func = $_[0];
67
68 unless (defined &$func) {
69 my $pkg = $func;
70 do {
71 $pkg =~ s/::[^:]+$//
72 or return;
73 eval "require $pkg";
74 } until defined &$func;
75 }
76
77 \&$func
78}
79
80sub MAX_READ_SIZE() { 131072 }
81
61=head1 METHODS 82=head1 METHODS
62 83
63=over 4 84=over 4
64 85
65=item B<new (%args)> 86=item $handle = B<new> AnyEvent::Handle fh => $filehandle, key => value...
66 87
67The constructor supports these arguments (all as key => value pairs). 88The constructor supports these arguments (all as C<< key => value >> pairs).
68 89
69=over 4 90=over 4
70 91
71=item fh => $filehandle [MANDATORY] 92=item fh => $filehandle [C<fh> or C<connect> MANDATORY]
72 93
73The filehandle this L<AnyEvent::Handle> object will operate on. 94The filehandle this L<AnyEvent::Handle> object will operate on.
74
75NOTE: The filehandle will be set to non-blocking (using 95NOTE: The filehandle will be set to non-blocking mode (using
76AnyEvent::Util::fh_nonblocking). 96C<AnyEvent::Util::fh_nonblocking>) by the constructor and needs to stay in
97that mode.
77 98
99=item connect => [$host, $service] [C<fh> or C<connect> MANDATORY]
100
101Try to connect to the specified host and service (port), using
102C<AnyEvent::Socket::tcp_connect>. The C<$host> additionally becomes the
103default C<peername>.
104
105You have to specify either this parameter, or C<fh>, above.
106
107It is possible to push requests on the read and write queues, and modify
108properties of the stream, even while AnyEvent::Handle is connecting.
109
110When this parameter is specified, then the C<on_prepare>,
111C<on_connect_error> and C<on_connect> callbacks will be called under the
112appropriate circumstances:
113
114=over 4
115
78=item on_eof => $cb->($handle) 116=item on_prepare => $cb->($handle)
79 117
80Set the callback to be called when an end-of-file condition is detected, 118This (rarely used) callback is called before a new connection is
81i.e. in the case of a socket, when the other side has closed the 119attempted, but after the file handle has been created. It could be used to
82connection cleanly. 120prepare the file handle with parameters required for the actual connect
121(as opposed to settings that can be changed when the connection is already
122established).
83 123
84While not mandatory, it is highly recommended to set an eof callback, 124The return value of this callback should be the connect timeout value in
85otherwise you might end up with a closed socket while you are still 125seconds (or C<0>, or C<undef>, or the empty list, to indicate that the
86waiting for data. 126default timeout is to be used).
87 127
128=item on_connect => $cb->($handle, $host, $port, $retry->())
129
130This callback is called when a connection has been successfully established.
131
132The peer's numeric host and port (the socket peername) are passed as
133parameters, together with a retry callback.
134
135If, for some reason, the handle is not acceptable, calling C<$retry>
136will continue with the next connection target (in case of multi-homed
137hosts or SRV records there can be multiple connection endpoints). At the
138time it is called the read and write queues, eof status, tls status and
139similar properties of the handle will have been reset.
140
141In most cases, you should ignore the C<$retry> parameter.
142
143=item on_connect_error => $cb->($handle, $message)
144
145This callback is called when the connection could not be
146established. C<$!> will contain the relevant error code, and C<$message> a
147message describing it (usually the same as C<"$!">).
148
149If this callback isn't specified, then C<on_error> will be called with a
150fatal error instead.
151
152=back
153
88=item on_error => $cb->($handle, $fatal) 154=item on_error => $cb->($handle, $fatal, $message)
89 155
90This is the error callback, which is called when, well, some error 156This is the error callback, which is called when, well, some error
91occured, such as not being able to resolve the hostname, failure to 157occured, such as not being able to resolve the hostname, failure to
92connect or a read error. 158connect, or a read error.
93 159
94Some errors are fatal (which is indicated by C<$fatal> being true). On 160Some errors are fatal (which is indicated by C<$fatal> being true). On
95fatal errors the handle object will be shut down and will not be 161fatal errors the handle object will be destroyed (by a call to C<< ->
162destroy >>) after invoking the error callback (which means you are free to
163examine the handle object). Examples of fatal errors are an EOF condition
164with active (but unsatisfiable) read watchers (C<EPIPE>) or I/O errors. In
165cases where the other side can close the connection at will, it is
166often easiest to not report C<EPIPE> errors in this callback.
167
168AnyEvent::Handle tries to find an appropriate error code for you to check
169against, but in some cases (TLS errors), this does not work well. It is
170recommended to always output the C<$message> argument in human-readable
171error messages (it's usually the same as C<"$!">).
172
96usable. Non-fatal errors can be retried by simply returning, but it is 173Non-fatal errors can be retried by returning, but it is recommended
97recommended to simply ignore this parameter and instead abondon the handle 174to simply ignore this parameter and instead abondon the handle object
98object when this callback is invoked. 175when this callback is invoked. Examples of non-fatal errors are timeouts
176C<ETIMEDOUT>) or badly-formatted data (C<EBADMSG>).
99 177
100On callback entrance, the value of C<$!> contains the operating system 178On entry to the callback, the value of C<$!> contains the operating
101error (or C<ENOSPC>, C<EPIPE>, C<ETIMEDOUT> or C<EBADMSG>). 179system error code (or C<ENOSPC>, C<EPIPE>, C<ETIMEDOUT>, C<EBADMSG> or
180C<EPROTO>).
102 181
103While not mandatory, it is I<highly> recommended to set this callback, as 182While not mandatory, it is I<highly> recommended to set this callback, as
104you will not be notified of errors otherwise. The default simply calls 183you will not be notified of errors otherwise. The default just calls
105C<croak>. 184C<croak>.
106 185
107=item on_read => $cb->($handle) 186=item on_read => $cb->($handle)
108 187
109This sets the default read callback, which is called when data arrives 188This sets the default read callback, which is called when data arrives
110and no read request is in the queue (unlike read queue callbacks, this 189and no read request is in the queue (unlike read queue callbacks, this
111callback will only be called when at least one octet of data is in the 190callback will only be called when at least one octet of data is in the
112read buffer). 191read buffer).
113 192
114To access (and remove data from) the read buffer, use the C<< ->rbuf >> 193To access (and remove data from) the read buffer, use the C<< ->rbuf >>
115method or access the C<$handle->{rbuf}> member directly. 194method or access the C<< $handle->{rbuf} >> member directly. Note that you
195must not enlarge or modify the read buffer, you can only remove data at
196the beginning from it.
116 197
198You can also call C<< ->push_read (...) >> or any other function that
199modifies the read queue. Or do both. Or ...
200
117When an EOF condition is detected then AnyEvent::Handle will first try to 201When an EOF condition is detected, AnyEvent::Handle will first try to
118feed all the remaining data to the queued callbacks and C<on_read> before 202feed all the remaining data to the queued callbacks and C<on_read> before
119calling the C<on_eof> callback. If no progress can be made, then a fatal 203calling the C<on_eof> callback. If no progress can be made, then a fatal
120error will be raised (with C<$!> set to C<EPIPE>). 204error will be raised (with C<$!> set to C<EPIPE>).
121 205
206Note that, unlike requests in the read queue, an C<on_read> callback
207doesn't mean you I<require> some data: if there is an EOF and there
208are outstanding read requests then an error will be flagged. With an
209C<on_read> callback, the C<on_eof> callback will be invoked.
210
211=item on_eof => $cb->($handle)
212
213Set the callback to be called when an end-of-file condition is detected,
214i.e. in the case of a socket, when the other side has closed the
215connection cleanly, and there are no outstanding read requests in the
216queue (if there are read requests, then an EOF counts as an unexpected
217connection close and will be flagged as an error).
218
219For sockets, this just means that the other side has stopped sending data,
220you can still try to write data, and, in fact, one can return from the EOF
221callback and continue writing data, as only the read part has been shut
222down.
223
224If an EOF condition has been detected but no C<on_eof> callback has been
225set, then a fatal error will be raised with C<$!> set to <0>.
226
122=item on_drain => $cb->($handle) 227=item on_drain => $cb->($handle)
123 228
124This sets the callback that is called when the write buffer becomes empty 229This sets the callback that is called when the write buffer becomes empty
125(or when the callback is set and the buffer is empty already). 230(or immediately if the buffer is empty already).
126 231
127To append to the write buffer, use the C<< ->push_write >> method. 232To append to the write buffer, use the C<< ->push_write >> method.
128 233
129This callback is useful when you don't want to put all of your write data 234This callback is useful when you don't want to put all of your write data
130into the queue at once, for example, when you want to write the contents 235into the queue at once, for example, when you want to write the contents
132memory and push it into the queue, but instead only read more data from 237memory and push it into the queue, but instead only read more data from
133the file when the write queue becomes empty. 238the file when the write queue becomes empty.
134 239
135=item timeout => $fractional_seconds 240=item timeout => $fractional_seconds
136 241
242=item rtimeout => $fractional_seconds
243
244=item wtimeout => $fractional_seconds
245
137If non-zero, then this enables an "inactivity" timeout: whenever this many 246If non-zero, then these enables an "inactivity" timeout: whenever this
138seconds pass without a successful read or write on the underlying file 247many seconds pass without a successful read or write on the underlying
139handle, the C<on_timeout> callback will be invoked (and if that one is 248file handle (or a call to C<timeout_reset>), the C<on_timeout> callback
140missing, an C<ETIMEDOUT> error will be raised). 249will be invoked (and if that one is missing, a non-fatal C<ETIMEDOUT>
250error will be raised).
141 251
252There are three variants of the timeouts that work independently
253of each other, for both read and write, just read, and just write:
254C<timeout>, C<rtimeout> and C<wtimeout>, with corresponding callbacks
255C<on_timeout>, C<on_rtimeout> and C<on_wtimeout>, and reset functions
256C<timeout_reset>, C<rtimeout_reset>, and C<wtimeout_reset>.
257
142Note that timeout processing is also active when you currently do not have 258Note that timeout processing is active even when you do not have
143any outstanding read or write requests: If you plan to keep the connection 259any outstanding read or write requests: If you plan to keep the connection
144idle then you should disable the timout temporarily or ignore the timeout 260idle then you should disable the timeout temporarily or ignore the timeout
145in the C<on_timeout> callback. 261in the C<on_timeout> callback, in which case AnyEvent::Handle will simply
262restart the timeout.
146 263
147Zero (the default) disables this timeout. 264Zero (the default) disables this timeout.
148 265
149=item on_timeout => $cb->($handle) 266=item on_timeout => $cb->($handle)
150 267
154 271
155=item rbuf_max => <bytes> 272=item rbuf_max => <bytes>
156 273
157If defined, then a fatal error will be raised (with C<$!> set to C<ENOSPC>) 274If defined, then a fatal error will be raised (with C<$!> set to C<ENOSPC>)
158when the read buffer ever (strictly) exceeds this size. This is useful to 275when the read buffer ever (strictly) exceeds this size. This is useful to
159avoid denial-of-service attacks. 276avoid some forms of denial-of-service attacks.
160 277
161For example, a server accepting connections from untrusted sources should 278For example, a server accepting connections from untrusted sources should
162be configured to accept only so-and-so much data that it cannot act on 279be configured to accept only so-and-so much data that it cannot act on
163(for example, when expecting a line, an attacker could send an unlimited 280(for example, when expecting a line, an attacker could send an unlimited
164amount of data without a callback ever being called as long as the line 281amount of data without a callback ever being called as long as the line
165isn't finished). 282isn't finished).
166 283
167=item autocork => <boolean> 284=item autocork => <boolean>
168 285
169When disabled (the default), then C<push_write> will try to immediately 286When disabled (the default), C<push_write> will try to immediately
170write the data to the handle if possible. This avoids having to register 287write the data to the handle if possible. This avoids having to register
171a write watcher and wait for the next event loop iteration, but can be 288a write watcher and wait for the next event loop iteration, but can
172inefficient if you write multiple small chunks (this disadvantage is 289be inefficient if you write multiple small chunks (on the wire, this
173usually avoided by your kernel's nagle algorithm, see C<low_delay>). 290disadvantage is usually avoided by your kernel's nagle algorithm, see
291C<no_delay>, but this option can save costly syscalls).
174 292
175When enabled, then writes will always be queued till the next event loop 293When enabled, writes will always be queued till the next event loop
176iteration. This is efficient when you do many small writes per iteration, 294iteration. This is efficient when you do many small writes per iteration,
177but less efficient when you do a single write only. 295but less efficient when you do a single write only per iteration (or when
296the write buffer often is full). It also increases write latency.
178 297
179=item no_delay => <boolean> 298=item no_delay => <boolean>
180 299
181When doing small writes on sockets, your operating system kernel might 300When doing small writes on sockets, your operating system kernel might
182wait a bit for more data before actually sending it out. This is called 301wait a bit for more data before actually sending it out. This is called
183the Nagle algorithm, and usually it is beneficial. 302the Nagle algorithm, and usually it is beneficial.
184 303
185In some situations you want as low a delay as possible, which cna be 304In some situations you want as low a delay as possible, which can be
186accomplishd by setting this option to true. 305accomplishd by setting this option to a true value.
187 306
188The default is your opertaing system's default behaviour, this option 307The default is your operating system's default behaviour (most likely
189explicitly enables or disables it, if possible. 308enabled). This option explicitly enables or disables it, if possible.
309
310=item keepalive => <boolean>
311
312Enables (default disable) the SO_KEEPALIVE option on the stream socket:
313normally, TCP connections have no time-out once established, so TCP
314connections, once established, can stay alive forever even when the other
315side has long gone. TCP keepalives are a cheap way to take down long-lived
316TCP connections when the other side becomes unreachable. While the default
317is OS-dependent, TCP keepalives usually kick in after around two hours,
318and, if the other side doesn't reply, take down the TCP connection some 10
319to 15 minutes later.
320
321It is harmless to specify this option for file handles that do not support
322keepalives, and enabling it on connections that are potentially long-lived
323is usually a good idea.
324
325=item oobinline => <boolean>
326
327BSD majorly fucked up the implementation of TCP urgent data. The result
328is that almost no OS implements TCP according to the specs, and every OS
329implements it slightly differently.
330
331If you want to handle TCP urgent data, then setting this flag (the default
332is enabled) gives you the most portable way of getting urgent data, by
333putting it into the stream.
334
335Since BSD emulation of OOB data on top of TCP's urgent data can have
336security implications, AnyEvent::Handle sets this flag automatically
337unless explicitly specified. Note that setting this flag after
338establishing a connection I<may> be a bit too late (data loss could
339already have occured on BSD systems), but at least it will protect you
340from most attacks.
190 341
191=item read_size => <bytes> 342=item read_size => <bytes>
192 343
193The default read block size (the amount of bytes this module will try to read 344The initial read block size, the number of bytes this module will try to
194during each (loop iteration). Default: C<8192>. 345read during each loop iteration. Each handle object will consume at least
346this amount of memory for the read buffer as well, so when handling many
347connections requirements). See also C<max_read_size>. Default: C<2048>.
348
349=item max_read_size => <bytes>
350
351The maximum read buffer size used by the dynamic adjustment
352algorithm: Each time AnyEvent::Handle can read C<read_size> bytes in
353one go it will double C<read_size> up to the maximum given by this
354option. Default: C<131072> or C<read_size>, whichever is higher.
195 355
196=item low_water_mark => <bytes> 356=item low_water_mark => <bytes>
197 357
198Sets the amount of bytes (default: C<0>) that make up an "empty" write 358Sets the number of bytes (default: C<0>) that make up an "empty" write
199buffer: If the write reaches this size or gets even samller it is 359buffer: If the buffer reaches this size or gets even samller it is
200considered empty. 360considered empty.
201 361
362Sometimes it can be beneficial (for performance reasons) to add data to
363the write buffer before it is fully drained, but this is a rare case, as
364the operating system kernel usually buffers data as well, so the default
365is good in almost all cases.
366
202=item linger => <seconds> 367=item linger => <seconds>
203 368
204If non-zero (default: C<3600>), then the destructor of the 369If this is non-zero (default: C<3600>), the destructor of the
205AnyEvent::Handle object will check wether there is still outstanding write 370AnyEvent::Handle object will check whether there is still outstanding
206data and will install a watcher that will write out this data. No errors 371write data and will install a watcher that will write this data to the
207will be reported (this mostly matches how the operating system treats 372socket. No errors will be reported (this mostly matches how the operating
208outstanding data at socket close time). 373system treats outstanding data at socket close time).
209 374
210This will not work for partial TLS data that could not yet been 375This will not work for partial TLS data that could not be encoded
211encoded. This data will be lost. 376yet. This data will be lost. Calling the C<stoptls> method in time might
377help.
378
379=item peername => $string
380
381A string used to identify the remote site - usually the DNS hostname
382(I<not> IDN!) used to create the connection, rarely the IP address.
383
384Apart from being useful in error messages, this string is also used in TLS
385peername verification (see C<verify_peername> in L<AnyEvent::TLS>). This
386verification will be skipped when C<peername> is not specified or is
387C<undef>.
212 388
213=item tls => "accept" | "connect" | Net::SSLeay::SSL object 389=item tls => "accept" | "connect" | Net::SSLeay::SSL object
214 390
215When this parameter is given, it enables TLS (SSL) mode, that means it 391When this parameter is given, it enables TLS (SSL) mode, that means
216will start making tls handshake and will transparently encrypt/decrypt 392AnyEvent will start a TLS handshake as soon as the connection has been
217data. 393established and will transparently encrypt/decrypt data afterwards.
394
395All TLS protocol errors will be signalled as C<EPROTO>, with an
396appropriate error message.
218 397
219TLS mode requires Net::SSLeay to be installed (it will be loaded 398TLS mode requires Net::SSLeay to be installed (it will be loaded
220automatically when you try to create a TLS handle). 399automatically when you try to create a TLS handle): this module doesn't
400have a dependency on that module, so if your module requires it, you have
401to add the dependency yourself.
221 402
222For the TLS server side, use C<accept>, and for the TLS client side of a 403Unlike TCP, TLS has a server and client side: for the TLS server side, use
223connection, use C<connect> mode. 404C<accept>, and for the TLS client side of a connection, use C<connect>
405mode.
224 406
225You can also provide your own TLS connection object, but you have 407You can also provide your own TLS connection object, but you have
226to make sure that you call either C<Net::SSLeay::set_connect_state> 408to make sure that you call either C<Net::SSLeay::set_connect_state>
227or C<Net::SSLeay::set_accept_state> on it before you pass it to 409or C<Net::SSLeay::set_accept_state> on it before you pass it to
228AnyEvent::Handle. 410AnyEvent::Handle. Also, this module will take ownership of this connection
411object.
229 412
413At some future point, AnyEvent::Handle might switch to another TLS
414implementation, then the option to use your own session object will go
415away.
416
417B<IMPORTANT:> since Net::SSLeay "objects" are really only integers,
418passing in the wrong integer will lead to certain crash. This most often
419happens when one uses a stylish C<< tls => 1 >> and is surprised about the
420segmentation fault.
421
230See the C<starttls> method if you need to start TLS negotiation later. 422Use the C<< ->starttls >> method if you need to start TLS negotiation later.
231 423
232=item tls_ctx => $ssl_ctx 424=item tls_ctx => $anyevent_tls
233 425
234Use the given Net::SSLeay::CTX object to create the new TLS connection 426Use the given C<AnyEvent::TLS> object to create the new TLS connection
235(unless a connection object was specified directly). If this parameter is 427(unless a connection object was specified directly). If this parameter is
236missing, then AnyEvent::Handle will use C<AnyEvent::Handle::TLS_CTX>. 428missing, then AnyEvent::Handle will use C<AnyEvent::Handle::TLS_CTX>.
237 429
430Instead of an object, you can also specify a hash reference with C<< key
431=> value >> pairs. Those will be passed to L<AnyEvent::TLS> to create a
432new TLS context object.
433
434=item on_starttls => $cb->($handle, $success[, $error_message])
435
436This callback will be invoked when the TLS/SSL handshake has finished. If
437C<$success> is true, then the TLS handshake succeeded, otherwise it failed
438(C<on_stoptls> will not be called in this case).
439
440The session in C<< $handle->{tls} >> can still be examined in this
441callback, even when the handshake was not successful.
442
443TLS handshake failures will not cause C<on_error> to be invoked when this
444callback is in effect, instead, the error message will be passed to C<on_starttls>.
445
446Without this callback, handshake failures lead to C<on_error> being
447called as usual.
448
449Note that you cannot just call C<starttls> again in this callback. If you
450need to do that, start an zero-second timer instead whose callback can
451then call C<< ->starttls >> again.
452
453=item on_stoptls => $cb->($handle)
454
455When a SSLv3/TLS shutdown/close notify/EOF is detected and this callback is
456set, then it will be invoked after freeing the TLS session. If it is not,
457then a TLS shutdown condition will be treated like a normal EOF condition
458on the handle.
459
460The session in C<< $handle->{tls} >> can still be examined in this
461callback.
462
463This callback will only be called on TLS shutdowns, not when the
464underlying handle signals EOF.
465
238=item json => JSON or JSON::XS object 466=item json => JSON or JSON::XS object
239 467
240This is the json coder object used by the C<json> read and write types. 468This is the json coder object used by the C<json> read and write types.
241 469
242If you don't supply it, then AnyEvent::Handle will create and use a 470If you don't supply it, then AnyEvent::Handle will create and use a
243suitable one, which will write and expect UTF-8 encoded JSON texts. 471suitable one (on demand), which will write and expect UTF-8 encoded JSON
472texts.
244 473
245Note that you are responsible to depend on the JSON module if you want to 474Note that you are responsible to depend on the JSON module if you want to
246use this functionality, as AnyEvent does not have a dependency itself. 475use this functionality, as AnyEvent does not have a dependency itself.
247 476
248=item filter_r => $cb
249
250=item filter_w => $cb
251
252These exist, but are undocumented at this time.
253
254=back 477=back
255 478
256=cut 479=cut
257 480
258sub new { 481sub new {
259 my $class = shift; 482 my $class = shift;
260
261 my $self = bless { @_ }, $class; 483 my $self = bless { @_ }, $class;
262 484
263 $self->{fh} or Carp::croak "mandatory argument fh is missing"; 485 if ($self->{fh}) {
486 $self->_start;
487 return unless $self->{fh}; # could be gone by now
488
489 } elsif ($self->{connect}) {
490 require AnyEvent::Socket;
491
492 $self->{peername} = $self->{connect}[0]
493 unless exists $self->{peername};
494
495 $self->{_skip_drain_rbuf} = 1;
496
497 {
498 Scalar::Util::weaken (my $self = $self);
499
500 $self->{_connect} =
501 AnyEvent::Socket::tcp_connect (
502 $self->{connect}[0],
503 $self->{connect}[1],
504 sub {
505 my ($fh, $host, $port, $retry) = @_;
506
507 delete $self->{_connect}; # no longer needed
508
509 if ($fh) {
510 $self->{fh} = $fh;
511
512 delete $self->{_skip_drain_rbuf};
513 $self->_start;
514
515 $self->{on_connect}
516 and $self->{on_connect}($self, $host, $port, sub {
517 delete @$self{qw(fh _tw _rtw _wtw _ww _rw _eof _queue rbuf _wbuf tls _tls_rbuf _tls_wbuf)};
518 $self->{_skip_drain_rbuf} = 1;
519 &$retry;
520 });
521
522 } else {
523 if ($self->{on_connect_error}) {
524 $self->{on_connect_error}($self, "$!");
525 $self->destroy;
526 } else {
527 $self->_error ($!, 1);
528 }
529 }
530 },
531 sub {
532 local $self->{fh} = $_[0];
533
534 $self->{on_prepare}
535 ? $self->{on_prepare}->($self)
536 : ()
537 }
538 );
539 }
540
541 } else {
542 Carp::croak "AnyEvent::Handle: either an existing fh or the connect parameter must be specified";
543 }
544
545 $self
546}
547
548sub _start {
549 my ($self) = @_;
550
551 # too many clueless people try to use udp and similar sockets
552 # with AnyEvent::Handle, do them a favour.
553 my $type = getsockopt $self->{fh}, Socket::SOL_SOCKET (), Socket::SO_TYPE ();
554 Carp::croak "AnyEvent::Handle: only stream sockets supported, anything else will NOT work!"
555 if Socket::SOCK_STREAM () != (unpack "I", $type) && defined $type;
264 556
265 AnyEvent::Util::fh_nonblocking $self->{fh}, 1; 557 AnyEvent::Util::fh_nonblocking $self->{fh}, 1;
266 558
267 if ($self->{tls}) { 559 $self->{_activity} =
268 require Net::SSLeay; 560 $self->{_ractivity} =
561 $self->{_wactivity} = AE::now;
562
563 $self->{read_size} ||= 2048;
564 $self->{max_read_size} = $self->{read_size}
565 if $self->{read_size} > ($self->{max_read_size} || MAX_READ_SIZE);
566
567 $self->timeout (delete $self->{timeout} ) if $self->{timeout};
568 $self->rtimeout (delete $self->{rtimeout} ) if $self->{rtimeout};
569 $self->wtimeout (delete $self->{wtimeout} ) if $self->{wtimeout};
570
571 $self->no_delay (delete $self->{no_delay} ) if exists $self->{no_delay} && $self->{no_delay};
572 $self->keepalive (delete $self->{keepalive}) if exists $self->{keepalive} && $self->{keepalive};
573
574 $self->oobinline (exists $self->{oobinline} ? delete $self->{oobinline} : 1);
575
269 $self->starttls (delete $self->{tls}, delete $self->{tls_ctx}); 576 $self->starttls (delete $self->{tls}, delete $self->{tls_ctx})
270 } 577 if $self->{tls};
271 578
272 $self->{_activity} = AnyEvent->now;
273 $self->_timeout;
274
275 $self->on_drain (delete $self->{on_drain}) if exists $self->{on_drain}; 579 $self->on_drain (delete $self->{on_drain} ) if $self->{on_drain};
276 $self->no_delay (delete $self->{no_delay}) if exists $self->{no_delay};
277 580
278 $self->start_read 581 $self->start_read
279 if $self->{on_read}; 582 if $self->{on_read} || @{ $self->{_queue} };
280 583
281 $self 584 $self->_drain_wbuf;
282}
283
284sub _shutdown {
285 my ($self) = @_;
286
287 delete $self->{_tw};
288 delete $self->{_rw};
289 delete $self->{_ww};
290 delete $self->{fh};
291
292 $self->stoptls;
293} 585}
294 586
295sub _error { 587sub _error {
296 my ($self, $errno, $fatal) = @_; 588 my ($self, $errno, $fatal, $message) = @_;
297
298 $self->_shutdown
299 if $fatal;
300 589
301 $! = $errno; 590 $! = $errno;
591 $message ||= "$!";
302 592
303 if ($self->{on_error}) { 593 if ($self->{on_error}) {
304 $self->{on_error}($self, $fatal); 594 $self->{on_error}($self, $fatal, $message);
305 } else { 595 $self->destroy if $fatal;
596 } elsif ($self->{fh} || $self->{connect}) {
597 $self->destroy;
306 Carp::croak "AnyEvent::Handle uncaught error: $!"; 598 Carp::croak "AnyEvent::Handle uncaught error: $message";
307 } 599 }
308} 600}
309 601
310=item $fh = $handle->fh 602=item $fh = $handle->fh
311 603
312This method returns the file handle of the L<AnyEvent::Handle> object. 604This method returns the file handle used to create the L<AnyEvent::Handle> object.
313 605
314=cut 606=cut
315 607
316sub fh { $_[0]{fh} } 608sub fh { $_[0]{fh} }
317 609
335 $_[0]{on_eof} = $_[1]; 627 $_[0]{on_eof} = $_[1];
336} 628}
337 629
338=item $handle->on_timeout ($cb) 630=item $handle->on_timeout ($cb)
339 631
340Replace the current C<on_timeout> callback, or disables the callback 632=item $handle->on_rtimeout ($cb)
341(but not the timeout) if C<$cb> = C<undef>. See C<timeout> constructor
342argument.
343 633
344=cut 634=item $handle->on_wtimeout ($cb)
345 635
346sub on_timeout { 636Replace the current C<on_timeout>, C<on_rtimeout> or C<on_wtimeout>
347 $_[0]{on_timeout} = $_[1]; 637callback, or disables the callback (but not the timeout) if C<$cb> =
348} 638C<undef>. See the C<timeout> constructor argument and method.
639
640=cut
641
642# see below
349 643
350=item $handle->autocork ($boolean) 644=item $handle->autocork ($boolean)
351 645
352Enables or disables the current autocork behaviour (see C<autocork> 646Enables or disables the current autocork behaviour (see C<autocork>
353constructor argument). 647constructor argument). Changes will only take effect on the next write.
354 648
355=cut 649=cut
650
651sub autocork {
652 $_[0]{autocork} = $_[1];
653}
356 654
357=item $handle->no_delay ($boolean) 655=item $handle->no_delay ($boolean)
358 656
359Enables or disables the C<no_delay> setting (see constructor argument of 657Enables or disables the C<no_delay> setting (see constructor argument of
360the same name for details). 658the same name for details).
362=cut 660=cut
363 661
364sub no_delay { 662sub no_delay {
365 $_[0]{no_delay} = $_[1]; 663 $_[0]{no_delay} = $_[1];
366 664
665 setsockopt $_[0]{fh}, Socket::IPPROTO_TCP (), Socket::TCP_NODELAY (), int $_[1]
666 if $_[0]{fh};
667}
668
669=item $handle->keepalive ($boolean)
670
671Enables or disables the C<keepalive> setting (see constructor argument of
672the same name for details).
673
674=cut
675
676sub keepalive {
677 $_[0]{keepalive} = $_[1];
678
367 eval { 679 eval {
368 local $SIG{__DIE__}; 680 local $SIG{__DIE__};
369 setsockopt $_[0]{fh}, &Socket::IPPROTO_TCP, &Socket::TCP_NODELAY, int $_[1]; 681 setsockopt $_[0]{fh}, Socket::SOL_SOCKET (), Socket::SO_KEEPALIVE (), int $_[1]
682 if $_[0]{fh};
370 }; 683 };
371} 684}
372 685
686=item $handle->oobinline ($boolean)
687
688Enables or disables the C<oobinline> setting (see constructor argument of
689the same name for details).
690
691=cut
692
693sub oobinline {
694 $_[0]{oobinline} = $_[1];
695
696 eval {
697 local $SIG{__DIE__};
698 setsockopt $_[0]{fh}, Socket::SOL_SOCKET (), Socket::SO_OOBINLINE (), int $_[1]
699 if $_[0]{fh};
700 };
701}
702
703=item $handle->keepalive ($boolean)
704
705Enables or disables the C<keepalive> setting (see constructor argument of
706the same name for details).
707
708=cut
709
710sub keepalive {
711 $_[0]{keepalive} = $_[1];
712
713 eval {
714 local $SIG{__DIE__};
715 setsockopt $_[0]{fh}, Socket::SOL_SOCKET (), Socket::SO_KEEPALIVE (), int $_[1]
716 if $_[0]{fh};
717 };
718}
719
720=item $handle->on_starttls ($cb)
721
722Replace the current C<on_starttls> callback (see the C<on_starttls> constructor argument).
723
724=cut
725
726sub on_starttls {
727 $_[0]{on_starttls} = $_[1];
728}
729
730=item $handle->on_stoptls ($cb)
731
732Replace the current C<on_stoptls> callback (see the C<on_stoptls> constructor argument).
733
734=cut
735
736sub on_stoptls {
737 $_[0]{on_stoptls} = $_[1];
738}
739
740=item $handle->rbuf_max ($max_octets)
741
742Configures the C<rbuf_max> setting (C<undef> disables it).
743
744=cut
745
746sub rbuf_max {
747 $_[0]{rbuf_max} = $_[1];
748}
749
373############################################################################# 750#############################################################################
374 751
375=item $handle->timeout ($seconds) 752=item $handle->timeout ($seconds)
376 753
754=item $handle->rtimeout ($seconds)
755
756=item $handle->wtimeout ($seconds)
757
377Configures (or disables) the inactivity timeout. 758Configures (or disables) the inactivity timeout.
378 759
379=cut 760=item $handle->timeout_reset
380 761
381sub timeout { 762=item $handle->rtimeout_reset
763
764=item $handle->wtimeout_reset
765
766Reset the activity timeout, as if data was received or sent.
767
768These methods are cheap to call.
769
770=cut
771
772for my $dir ("", "r", "w") {
773 my $timeout = "${dir}timeout";
774 my $tw = "_${dir}tw";
775 my $on_timeout = "on_${dir}timeout";
776 my $activity = "_${dir}activity";
777 my $cb;
778
779 *$on_timeout = sub {
780 $_[0]{$on_timeout} = $_[1];
781 };
782
783 *$timeout = sub {
382 my ($self, $timeout) = @_; 784 my ($self, $new_value) = @_;
383 785
786 $new_value >= 0
787 or Carp::croak "AnyEvent::Handle->$timeout called with negative timeout ($new_value), caught";
788
384 $self->{timeout} = $timeout; 789 $self->{$timeout} = $new_value;
385 $self->_timeout; 790 delete $self->{$tw}; &$cb;
386} 791 };
387 792
793 *{"${dir}timeout_reset"} = sub {
794 $_[0]{$activity} = AE::now;
795 };
796
797 # main workhorse:
388# reset the timeout watcher, as neccessary 798 # reset the timeout watcher, as neccessary
389# also check for time-outs 799 # also check for time-outs
390sub _timeout { 800 $cb = sub {
391 my ($self) = @_; 801 my ($self) = @_;
392 802
393 if ($self->{timeout}) { 803 if ($self->{$timeout} && $self->{fh}) {
394 my $NOW = AnyEvent->now; 804 my $NOW = AE::now;
395 805
396 # when would the timeout trigger? 806 # when would the timeout trigger?
397 my $after = $self->{_activity} + $self->{timeout} - $NOW; 807 my $after = $self->{$activity} + $self->{$timeout} - $NOW;
398 808
399 # now or in the past already? 809 # now or in the past already?
400 if ($after <= 0) { 810 if ($after <= 0) {
401 $self->{_activity} = $NOW; 811 $self->{$activity} = $NOW;
402 812
403 if ($self->{on_timeout}) { 813 if ($self->{$on_timeout}) {
404 $self->{on_timeout}($self); 814 $self->{$on_timeout}($self);
405 } else { 815 } else {
406 $self->_error (&Errno::ETIMEDOUT); 816 $self->_error (Errno::ETIMEDOUT);
817 }
818
819 # callback could have changed timeout value, optimise
820 return unless $self->{$timeout};
821
822 # calculate new after
823 $after = $self->{$timeout};
407 } 824 }
408 825
409 # callback could have changed timeout value, optimise 826 Scalar::Util::weaken $self;
410 return unless $self->{timeout}; 827 return unless $self; # ->error could have destroyed $self
411 828
412 # calculate new after 829 $self->{$tw} ||= AE::timer $after, 0, sub {
413 $after = $self->{timeout}; 830 delete $self->{$tw};
831 $cb->($self);
832 };
833 } else {
834 delete $self->{$tw};
414 } 835 }
415
416 Scalar::Util::weaken $self;
417 return unless $self; # ->error could have destroyed $self
418
419 $self->{_tw} ||= AnyEvent->timer (after => $after, cb => sub {
420 delete $self->{_tw};
421 $self->_timeout;
422 });
423 } else {
424 delete $self->{_tw};
425 } 836 }
426} 837}
427 838
428############################################################################# 839#############################################################################
429 840
445=item $handle->on_drain ($cb) 856=item $handle->on_drain ($cb)
446 857
447Sets the C<on_drain> callback or clears it (see the description of 858Sets the C<on_drain> callback or clears it (see the description of
448C<on_drain> in the constructor). 859C<on_drain> in the constructor).
449 860
861This method may invoke callbacks (and therefore the handle might be
862destroyed after it returns).
863
450=cut 864=cut
451 865
452sub on_drain { 866sub on_drain {
453 my ($self, $cb) = @_; 867 my ($self, $cb) = @_;
454 868
455 $self->{on_drain} = $cb; 869 $self->{on_drain} = $cb;
456 870
457 $cb->($self) 871 $cb->($self)
458 if $cb && $self->{low_water_mark} >= length $self->{wbuf}; 872 if $cb && $self->{low_water_mark} >= (length $self->{wbuf}) + (length $self->{_tls_wbuf});
459} 873}
460 874
461=item $handle->push_write ($data) 875=item $handle->push_write ($data)
462 876
463Queues the given scalar to be written. You can push as much data as you 877Queues the given scalar to be written. You can push as much data as you
464want (only limited by the available memory), as C<AnyEvent::Handle> 878want (only limited by the available memory), as C<AnyEvent::Handle>
465buffers it independently of the kernel. 879buffers it independently of the kernel.
466 880
881This method may invoke callbacks (and therefore the handle might be
882destroyed after it returns).
883
467=cut 884=cut
468 885
469sub _drain_wbuf { 886sub _drain_wbuf {
470 my ($self) = @_; 887 my ($self) = @_;
471 888
474 Scalar::Util::weaken $self; 891 Scalar::Util::weaken $self;
475 892
476 my $cb = sub { 893 my $cb = sub {
477 my $len = syswrite $self->{fh}, $self->{wbuf}; 894 my $len = syswrite $self->{fh}, $self->{wbuf};
478 895
479 if ($len >= 0) { 896 if (defined $len) {
480 substr $self->{wbuf}, 0, $len, ""; 897 substr $self->{wbuf}, 0, $len, "";
481 898
482 $self->{_activity} = AnyEvent->now; 899 $self->{_activity} = $self->{_wactivity} = AE::now;
483 900
484 $self->{on_drain}($self) 901 $self->{on_drain}($self)
485 if $self->{low_water_mark} >= length $self->{wbuf} 902 if $self->{low_water_mark} >= (length $self->{wbuf}) + (length $self->{_tls_wbuf})
486 && $self->{on_drain}; 903 && $self->{on_drain};
487 904
488 delete $self->{_ww} unless length $self->{wbuf}; 905 delete $self->{_ww} unless length $self->{wbuf};
489 } elsif ($! != EAGAIN && $! != EINTR && $! != WSAEWOULDBLOCK) { 906 } elsif ($! != EAGAIN && $! != EINTR && $! != WSAEWOULDBLOCK) {
490 $self->_error ($!, 1); 907 $self->_error ($!, 1);
493 910
494 # try to write data immediately 911 # try to write data immediately
495 $cb->() unless $self->{autocork}; 912 $cb->() unless $self->{autocork};
496 913
497 # if still data left in wbuf, we need to poll 914 # if still data left in wbuf, we need to poll
498 $self->{_ww} = AnyEvent->io (fh => $self->{fh}, poll => "w", cb => $cb) 915 $self->{_ww} = AE::io $self->{fh}, 1, $cb
499 if length $self->{wbuf}; 916 if length $self->{wbuf};
500 }; 917 };
501} 918}
502 919
503our %WH; 920our %WH;
504 921
922# deprecated
505sub register_write_type($$) { 923sub register_write_type($$) {
506 $WH{$_[0]} = $_[1]; 924 $WH{$_[0]} = $_[1];
507} 925}
508 926
509sub push_write { 927sub push_write {
510 my $self = shift; 928 my $self = shift;
511 929
512 if (@_ > 1) { 930 if (@_ > 1) {
513 my $type = shift; 931 my $type = shift;
514 932
933 @_ = ($WH{$type} ||= _load_func "$type\::anyevent_write_type"
515 @_ = ($WH{$type} or Carp::croak "unsupported type passed to AnyEvent::Handle::push_write") 934 or Carp::croak "unsupported/unloadable type '$type' passed to AnyEvent::Handle::push_write")
516 ->($self, @_); 935 ->($self, @_);
517 } 936 }
518 937
938 # we downgrade here to avoid hard-to-track-down bugs,
939 # and diagnose the problem earlier and better.
940
519 if ($self->{filter_w}) { 941 if ($self->{tls}) {
520 $self->{filter_w}($self, \$_[0]); 942 utf8::downgrade $self->{_tls_wbuf} .= $_[0];
943 &_dotls ($self) if $self->{fh};
521 } else { 944 } else {
522 $self->{wbuf} .= $_[0]; 945 utf8::downgrade $self->{wbuf} .= $_[0];
523 $self->_drain_wbuf; 946 $self->_drain_wbuf if $self->{fh};
524 } 947 }
525} 948}
526 949
527=item $handle->push_write (type => @args) 950=item $handle->push_write (type => @args)
528 951
529Instead of formatting your data yourself, you can also let this module do 952Instead of formatting your data yourself, you can also let this module
530the job by specifying a type and type-specific arguments. 953do the job by specifying a type and type-specific arguments. You
954can also specify the (fully qualified) name of a package, in which
955case AnyEvent tries to load the package and then expects to find the
956C<anyevent_write_type> function inside (see "custom write types", below).
531 957
532Predefined types are (if you have ideas for additional types, feel free to 958Predefined types are (if you have ideas for additional types, feel free to
533drop by and tell us): 959drop by and tell us):
534 960
535=over 4 961=over 4
542=cut 968=cut
543 969
544register_write_type netstring => sub { 970register_write_type netstring => sub {
545 my ($self, $string) = @_; 971 my ($self, $string) = @_;
546 972
547 sprintf "%d:%s,", (length $string), $string 973 (length $string) . ":$string,"
548}; 974};
549 975
550=item packstring => $format, $data 976=item packstring => $format, $data
551 977
552An octet string prefixed with an encoded length. The encoding C<$format> 978An octet string prefixed with an encoded length. The encoding C<$format>
592Other languages could read single lines terminated by a newline and pass 1018Other languages could read single lines terminated by a newline and pass
593this line into their JSON decoder of choice. 1019this line into their JSON decoder of choice.
594 1020
595=cut 1021=cut
596 1022
1023sub json_coder() {
1024 eval { require JSON::XS; JSON::XS->new->utf8 }
1025 || do { require JSON; JSON->new->utf8 }
1026}
1027
597register_write_type json => sub { 1028register_write_type json => sub {
598 my ($self, $ref) = @_; 1029 my ($self, $ref) = @_;
599 1030
600 require JSON; 1031 my $json = $self->{json} ||= json_coder;
601 1032
602 $self->{json} ? $self->{json}->encode ($ref) 1033 $json->encode ($ref)
603 : JSON::encode_json ($ref)
604}; 1034};
605 1035
606=item storable => $reference 1036=item storable => $reference
607 1037
608Freezes the given reference using L<Storable> and writes it to the 1038Freezes the given reference using L<Storable> and writes it to the
618 pack "w/a*", Storable::nfreeze ($ref) 1048 pack "w/a*", Storable::nfreeze ($ref)
619}; 1049};
620 1050
621=back 1051=back
622 1052
623=item AnyEvent::Handle::register_write_type type => $coderef->($handle, @args) 1053=item $handle->push_shutdown
624 1054
625This function (not method) lets you add your own types to C<push_write>. 1055Sometimes you know you want to close the socket after writing your data
1056before it was actually written. One way to do that is to replace your
1057C<on_drain> handler by a callback that shuts down the socket (and set
1058C<low_water_mark> to C<0>). This method is a shorthand for just that, and
1059replaces the C<on_drain> callback with:
1060
1061 sub { shutdown $_[0]{fh}, 1 } # for push_shutdown
1062
1063This simply shuts down the write side and signals an EOF condition to the
1064the peer.
1065
1066You can rely on the normal read queue and C<on_eof> handling
1067afterwards. This is the cleanest way to close a connection.
1068
1069This method may invoke callbacks (and therefore the handle might be
1070destroyed after it returns).
1071
1072=cut
1073
1074sub push_shutdown {
1075 my ($self) = @_;
1076
1077 delete $self->{low_water_mark};
1078 $self->on_drain (sub { shutdown $_[0]{fh}, 1 });
1079}
1080
1081=item custom write types - Package::anyevent_write_type $handle, @args
1082
1083Instead of one of the predefined types, you can also specify the name of
1084a package. AnyEvent will try to load the package and then expects to find
1085a function named C<anyevent_write_type> inside. If it isn't found, it
1086progressively tries to load the parent package until it either finds the
1087function (good) or runs out of packages (bad).
1088
626Whenever the given C<type> is used, C<push_write> will invoke the code 1089Whenever the given C<type> is used, C<push_write> will the function with
627reference with the handle object and the remaining arguments. 1090the handle object and the remaining arguments.
628 1091
629The code reference is supposed to return a single octet string that will 1092The function is supposed to return a single octet string that will be
630be appended to the write buffer. 1093appended to the write buffer, so you cna mentally treat this function as a
1094"arguments to on-the-wire-format" converter.
631 1095
632Note that this is a function, and all types registered this way will be 1096Example: implement a custom write type C<join> that joins the remaining
633global, so try to use unique names. 1097arguments using the first one.
1098
1099 $handle->push_write (My::Type => " ", 1,2,3);
1100
1101 # uses the following package, which can be defined in the "My::Type" or in
1102 # the "My" modules to be auto-loaded, or just about anywhere when the
1103 # My::Type::anyevent_write_type is defined before invoking it.
1104
1105 package My::Type;
1106
1107 sub anyevent_write_type {
1108 my ($handle, $delim, @args) = @_;
1109
1110 join $delim, @args
1111 }
634 1112
635=cut 1113=cut
636 1114
637############################################################################# 1115#############################################################################
638 1116
647ways, the "simple" way, using only C<on_read> and the "complex" way, using 1125ways, the "simple" way, using only C<on_read> and the "complex" way, using
648a queue. 1126a queue.
649 1127
650In the simple case, you just install an C<on_read> callback and whenever 1128In the simple case, you just install an C<on_read> callback and whenever
651new data arrives, it will be called. You can then remove some data (if 1129new data arrives, it will be called. You can then remove some data (if
652enough is there) from the read buffer (C<< $handle->rbuf >>). Or you cna 1130enough is there) from the read buffer (C<< $handle->rbuf >>). Or you can
653leave the data there if you want to accumulate more (e.g. when only a 1131leave the data there if you want to accumulate more (e.g. when only a
654partial message has been received so far). 1132partial message has been received so far), or change the read queue with
1133e.g. C<push_read>.
655 1134
656In the more complex case, you want to queue multiple callbacks. In this 1135In the more complex case, you want to queue multiple callbacks. In this
657case, AnyEvent::Handle will call the first queued callback each time new 1136case, AnyEvent::Handle will call the first queued callback each time new
658data arrives (also the first time it is queued) and removes it when it has 1137data arrives (also the first time it is queued) and remove it when it has
659done its job (see C<push_read>, below). 1138done its job (see C<push_read>, below).
660 1139
661This way you can, for example, push three line-reads, followed by reading 1140This way you can, for example, push three line-reads, followed by reading
662a chunk of data, and AnyEvent::Handle will execute them in order. 1141a chunk of data, and AnyEvent::Handle will execute them in order.
663 1142
720=cut 1199=cut
721 1200
722sub _drain_rbuf { 1201sub _drain_rbuf {
723 my ($self) = @_; 1202 my ($self) = @_;
724 1203
1204 # avoid recursion
1205 return if $self->{_skip_drain_rbuf};
725 local $self->{_in_drain} = 1; 1206 local $self->{_skip_drain_rbuf} = 1;
726
727 if (
728 defined $self->{rbuf_max}
729 && $self->{rbuf_max} < length $self->{rbuf}
730 ) {
731 return $self->_error (&Errno::ENOSPC, 1);
732 }
733 1207
734 while () { 1208 while () {
1209 # we need to use a separate tls read buffer, as we must not receive data while
1210 # we are draining the buffer, and this can only happen with TLS.
1211 $self->{rbuf} .= delete $self->{_tls_rbuf}
1212 if exists $self->{_tls_rbuf};
1213
735 my $len = length $self->{rbuf}; 1214 my $len = length $self->{rbuf};
736 1215
737 if (my $cb = shift @{ $self->{_queue} }) { 1216 if (my $cb = shift @{ $self->{_queue} }) {
738 unless ($cb->($self)) { 1217 unless ($cb->($self)) {
739 if ($self->{_eof}) { 1218 # no progress can be made
740 # no progress can be made (not enough data and no data forthcoming) 1219 # (not enough data and no data forthcoming)
741 $self->_error (&Errno::EPIPE, 1), last; 1220 $self->_error (Errno::EPIPE, 1), return
742 } 1221 if $self->{_eof};
743 1222
744 unshift @{ $self->{_queue} }, $cb; 1223 unshift @{ $self->{_queue} }, $cb;
745 last; 1224 last;
746 } 1225 }
747 } elsif ($self->{on_read}) { 1226 } elsif ($self->{on_read}) {
754 && !@{ $self->{_queue} } # and the queue is still empty 1233 && !@{ $self->{_queue} } # and the queue is still empty
755 && $self->{on_read} # but we still have on_read 1234 && $self->{on_read} # but we still have on_read
756 ) { 1235 ) {
757 # no further data will arrive 1236 # no further data will arrive
758 # so no progress can be made 1237 # so no progress can be made
759 $self->_error (&Errno::EPIPE, 1), last 1238 $self->_error (Errno::EPIPE, 1), return
760 if $self->{_eof}; 1239 if $self->{_eof};
761 1240
762 last; # more data might arrive 1241 last; # more data might arrive
763 } 1242 }
764 } else { 1243 } else {
765 # read side becomes idle 1244 # read side becomes idle
766 delete $self->{_rw}; 1245 delete $self->{_rw} unless $self->{tls};
767 last; 1246 last;
768 } 1247 }
769 } 1248 }
770 1249
1250 if ($self->{_eof}) {
1251 $self->{on_eof}
771 $self->{on_eof}($self) 1252 ? $self->{on_eof}($self)
772 if $self->{_eof} && $self->{on_eof}; 1253 : $self->_error (0, 1, "Unexpected end-of-file");
1254
1255 return;
1256 }
1257
1258 if (
1259 defined $self->{rbuf_max}
1260 && $self->{rbuf_max} < length $self->{rbuf}
1261 ) {
1262 $self->_error (Errno::ENOSPC, 1), return;
1263 }
773 1264
774 # may need to restart read watcher 1265 # may need to restart read watcher
775 unless ($self->{_rw}) { 1266 unless ($self->{_rw}) {
776 $self->start_read 1267 $self->start_read
777 if $self->{on_read} || @{ $self->{_queue} }; 1268 if $self->{on_read} || @{ $self->{_queue} };
782 1273
783This replaces the currently set C<on_read> callback, or clears it (when 1274This replaces the currently set C<on_read> callback, or clears it (when
784the new callback is C<undef>). See the description of C<on_read> in the 1275the new callback is C<undef>). See the description of C<on_read> in the
785constructor. 1276constructor.
786 1277
1278This method may invoke callbacks (and therefore the handle might be
1279destroyed after it returns).
1280
787=cut 1281=cut
788 1282
789sub on_read { 1283sub on_read {
790 my ($self, $cb) = @_; 1284 my ($self, $cb) = @_;
791 1285
792 $self->{on_read} = $cb; 1286 $self->{on_read} = $cb;
793 $self->_drain_rbuf if $cb && !$self->{_in_drain}; 1287 $self->_drain_rbuf if $cb;
794} 1288}
795 1289
796=item $handle->rbuf 1290=item $handle->rbuf
797 1291
798Returns the read buffer (as a modifiable lvalue). 1292Returns the read buffer (as a modifiable lvalue). You can also access the
1293read buffer directly as the C<< ->{rbuf} >> member, if you want (this is
1294much faster, and no less clean).
799 1295
800You can access the read buffer directly as the C<< ->{rbuf} >> member, if 1296The only operation allowed on the read buffer (apart from looking at it)
801you want. 1297is removing data from its beginning. Otherwise modifying or appending to
1298it is not allowed and will lead to hard-to-track-down bugs.
802 1299
803NOTE: The read buffer should only be used or modified if the C<on_read>, 1300NOTE: The read buffer should only be used or modified in the C<on_read>
804C<push_read> or C<unshift_read> methods are used. The other read methods 1301callback or when C<push_read> or C<unshift_read> are used with a single
805automatically manage the read buffer. 1302callback (i.e. untyped). Typed C<push_read> and C<unshift_read> methods
1303will manage the read buffer on their own.
806 1304
807=cut 1305=cut
808 1306
809sub rbuf : lvalue { 1307sub rbuf : lvalue {
810 $_[0]{rbuf} 1308 $_[0]{rbuf}
827 1325
828If enough data was available, then the callback must remove all data it is 1326If enough data was available, then the callback must remove all data it is
829interested in (which can be none at all) and return a true value. After returning 1327interested in (which can be none at all) and return a true value. After returning
830true, it will be removed from the queue. 1328true, it will be removed from the queue.
831 1329
1330These methods may invoke callbacks (and therefore the handle might be
1331destroyed after it returns).
1332
832=cut 1333=cut
833 1334
834our %RH; 1335our %RH;
835 1336
836sub register_read_type($$) { 1337sub register_read_type($$) {
842 my $cb = pop; 1343 my $cb = pop;
843 1344
844 if (@_) { 1345 if (@_) {
845 my $type = shift; 1346 my $type = shift;
846 1347
1348 $cb = ($RH{$type} ||= _load_func "$type\::anyevent_read_type"
847 $cb = ($RH{$type} or Carp::croak "unsupported type passed to AnyEvent::Handle::push_read") 1349 or Carp::croak "unsupported/unloadable type '$type' passed to AnyEvent::Handle::push_read")
848 ->($self, $cb, @_); 1350 ->($self, $cb, @_);
849 } 1351 }
850 1352
851 push @{ $self->{_queue} }, $cb; 1353 push @{ $self->{_queue} }, $cb;
852 $self->_drain_rbuf unless $self->{_in_drain}; 1354 $self->_drain_rbuf;
853} 1355}
854 1356
855sub unshift_read { 1357sub unshift_read {
856 my $self = shift; 1358 my $self = shift;
857 my $cb = pop; 1359 my $cb = pop;
858 1360
859 if (@_) { 1361 if (@_) {
860 my $type = shift; 1362 my $type = shift;
861 1363
1364 $cb = ($RH{$type} ||= _load_func "$type\::anyevent_read_type"
862 $cb = ($RH{$type} or Carp::croak "unsupported type passed to AnyEvent::Handle::unshift_read") 1365 or Carp::croak "unsupported/unloadable type '$type' passed to AnyEvent::Handle::unshift_read")
863 ->($self, $cb, @_); 1366 ->($self, $cb, @_);
864 } 1367 }
865 1368
866
867 unshift @{ $self->{_queue} }, $cb; 1369 unshift @{ $self->{_queue} }, $cb;
868 $self->_drain_rbuf unless $self->{_in_drain}; 1370 $self->_drain_rbuf;
869} 1371}
870 1372
871=item $handle->push_read (type => @args, $cb) 1373=item $handle->push_read (type => @args, $cb)
872 1374
873=item $handle->unshift_read (type => @args, $cb) 1375=item $handle->unshift_read (type => @args, $cb)
874 1376
875Instead of providing a callback that parses the data itself you can chose 1377Instead of providing a callback that parses the data itself you can chose
876between a number of predefined parsing formats, for chunks of data, lines 1378between a number of predefined parsing formats, for chunks of data, lines
877etc. 1379etc. You can also specify the (fully qualified) name of a package, in
1380which case AnyEvent tries to load the package and then expects to find the
1381C<anyevent_read_type> function inside (see "custom read types", below).
878 1382
879Predefined types are (if you have ideas for additional types, feel free to 1383Predefined types are (if you have ideas for additional types, feel free to
880drop by and tell us): 1384drop by and tell us):
881 1385
882=over 4 1386=over 4
974the receive buffer when neither C<$accept> nor C<$reject> match, 1478the receive buffer when neither C<$accept> nor C<$reject> match,
975and everything preceding and including the match will be accepted 1479and everything preceding and including the match will be accepted
976unconditionally. This is useful to skip large amounts of data that you 1480unconditionally. This is useful to skip large amounts of data that you
977know cannot be matched, so that the C<$accept> or C<$reject> regex do not 1481know cannot be matched, so that the C<$accept> or C<$reject> regex do not
978have to start matching from the beginning. This is purely an optimisation 1482have to start matching from the beginning. This is purely an optimisation
979and is usually worth only when you expect more than a few kilobytes. 1483and is usually worth it only when you expect more than a few kilobytes.
980 1484
981Example: expect a http header, which ends at C<\015\012\015\012>. Since we 1485Example: expect a http header, which ends at C<\015\012\015\012>. Since we
982expect the header to be very large (it isn't in practise, but...), we use 1486expect the header to be very large (it isn't in practice, but...), we use
983a skip regex to skip initial portions. The skip regex is tricky in that 1487a skip regex to skip initial portions. The skip regex is tricky in that
984it only accepts something not ending in either \015 or \012, as these are 1488it only accepts something not ending in either \015 or \012, as these are
985required for the accept regex. 1489required for the accept regex.
986 1490
987 $handle->push_read (regex => 1491 $handle->push_read (regex =>
1006 return 1; 1510 return 1;
1007 } 1511 }
1008 1512
1009 # reject 1513 # reject
1010 if ($reject && $$rbuf =~ $reject) { 1514 if ($reject && $$rbuf =~ $reject) {
1011 $self->_error (&Errno::EBADMSG); 1515 $self->_error (Errno::EBADMSG);
1012 } 1516 }
1013 1517
1014 # skip 1518 # skip
1015 if ($skip && $$rbuf =~ $skip) { 1519 if ($skip && $$rbuf =~ $skip) {
1016 $data .= substr $$rbuf, 0, $+[0], ""; 1520 $data .= substr $$rbuf, 0, $+[0], "";
1032 my ($self, $cb) = @_; 1536 my ($self, $cb) = @_;
1033 1537
1034 sub { 1538 sub {
1035 unless ($_[0]{rbuf} =~ s/^(0|[1-9][0-9]*)://) { 1539 unless ($_[0]{rbuf} =~ s/^(0|[1-9][0-9]*)://) {
1036 if ($_[0]{rbuf} =~ /[^0-9]/) { 1540 if ($_[0]{rbuf} =~ /[^0-9]/) {
1037 $self->_error (&Errno::EBADMSG); 1541 $self->_error (Errno::EBADMSG);
1038 } 1542 }
1039 return; 1543 return;
1040 } 1544 }
1041 1545
1042 my $len = $1; 1546 my $len = $1;
1045 my $string = $_[1]; 1549 my $string = $_[1];
1046 $_[0]->unshift_read (chunk => 1, sub { 1550 $_[0]->unshift_read (chunk => 1, sub {
1047 if ($_[1] eq ",") { 1551 if ($_[1] eq ",") {
1048 $cb->($_[0], $string); 1552 $cb->($_[0], $string);
1049 } else { 1553 } else {
1050 $self->_error (&Errno::EBADMSG); 1554 $self->_error (Errno::EBADMSG);
1051 } 1555 }
1052 }); 1556 });
1053 }); 1557 });
1054 1558
1055 1 1559 1
1061An octet string prefixed with an encoded length. The encoding C<$format> 1565An octet string prefixed with an encoded length. The encoding C<$format>
1062uses the same format as a Perl C<pack> format, but must specify a single 1566uses the same format as a Perl C<pack> format, but must specify a single
1063integer only (only one of C<cCsSlLqQiInNvVjJw> is allowed, plus an 1567integer only (only one of C<cCsSlLqQiInNvVjJw> is allowed, plus an
1064optional C<!>, C<< < >> or C<< > >> modifier). 1568optional C<!>, C<< < >> or C<< > >> modifier).
1065 1569
1066DNS over TCP uses a prefix of C<n>, EPP uses a prefix of C<N>. 1570For example, DNS over TCP uses a prefix of C<n> (2 octet network order),
1571EPP uses a prefix of C<N> (4 octtes).
1067 1572
1068Example: read a block of data prefixed by its length in BER-encoded 1573Example: read a block of data prefixed by its length in BER-encoded
1069format (very efficient). 1574format (very efficient).
1070 1575
1071 $handle->push_read (packstring => "w", sub { 1576 $handle->push_read (packstring => "w", sub {
1101 } 1606 }
1102}; 1607};
1103 1608
1104=item json => $cb->($handle, $hash_or_arrayref) 1609=item json => $cb->($handle, $hash_or_arrayref)
1105 1610
1106Reads a JSON object or array, decodes it and passes it to the callback. 1611Reads a JSON object or array, decodes it and passes it to the
1612callback. When a parse error occurs, an C<EBADMSG> error will be raised.
1107 1613
1108If a C<json> object was passed to the constructor, then that will be used 1614If a C<json> object was passed to the constructor, then that will be used
1109for the final decode, otherwise it will create a JSON coder expecting UTF-8. 1615for the final decode, otherwise it will create a JSON coder expecting UTF-8.
1110 1616
1111This read type uses the incremental parser available with JSON version 1617This read type uses the incremental parser available with JSON version
1120=cut 1626=cut
1121 1627
1122register_read_type json => sub { 1628register_read_type json => sub {
1123 my ($self, $cb) = @_; 1629 my ($self, $cb) = @_;
1124 1630
1125 require JSON; 1631 my $json = $self->{json} ||= json_coder;
1126 1632
1127 my $data; 1633 my $data;
1128 my $rbuf = \$self->{rbuf}; 1634 my $rbuf = \$self->{rbuf};
1129 1635
1130 my $json = $self->{json} ||= JSON->new->utf8;
1131
1132 sub { 1636 sub {
1133 my $ref = $json->incr_parse ($self->{rbuf}); 1637 my $ref = eval { $json->incr_parse ($self->{rbuf}) };
1134 1638
1135 if ($ref) { 1639 if ($ref) {
1136 $self->{rbuf} = $json->incr_text; 1640 $self->{rbuf} = $json->incr_text;
1137 $json->incr_text = ""; 1641 $json->incr_text = "";
1138 $cb->($self, $ref); 1642 $cb->($self, $ref);
1139 1643
1140 1 1644 1
1645 } elsif ($@) {
1646 # error case
1647 $json->incr_skip;
1648
1649 $self->{rbuf} = $json->incr_text;
1650 $json->incr_text = "";
1651
1652 $self->_error (Errno::EBADMSG);
1653
1654 ()
1141 } else { 1655 } else {
1142 $self->{rbuf} = ""; 1656 $self->{rbuf} = "";
1657
1143 () 1658 ()
1144 } 1659 }
1145 } 1660 }
1146}; 1661};
1147 1662
1179 # read remaining chunk 1694 # read remaining chunk
1180 $_[0]->unshift_read (chunk => $len, sub { 1695 $_[0]->unshift_read (chunk => $len, sub {
1181 if (my $ref = eval { Storable::thaw ($_[1]) }) { 1696 if (my $ref = eval { Storable::thaw ($_[1]) }) {
1182 $cb->($_[0], $ref); 1697 $cb->($_[0], $ref);
1183 } else { 1698 } else {
1184 $self->_error (&Errno::EBADMSG); 1699 $self->_error (Errno::EBADMSG);
1185 } 1700 }
1186 }); 1701 });
1187 } 1702 }
1188 1703
1189 1 1704 1
1190 } 1705 }
1191}; 1706};
1192 1707
1193=back 1708=back
1194 1709
1195=item AnyEvent::Handle::register_read_type type => $coderef->($handle, $cb, @args) 1710=item custom read types - Package::anyevent_read_type $handle, $cb, @args
1196 1711
1197This function (not method) lets you add your own types to C<push_read>. 1712Instead of one of the predefined types, you can also specify the name
1713of a package. AnyEvent will try to load the package and then expects to
1714find a function named C<anyevent_read_type> inside. If it isn't found, it
1715progressively tries to load the parent package until it either finds the
1716function (good) or runs out of packages (bad).
1198 1717
1199Whenever the given C<type> is used, C<push_read> will invoke the code 1718Whenever this type is used, C<push_read> will invoke the function with the
1200reference with the handle object, the callback and the remaining 1719handle object, the original callback and the remaining arguments.
1201arguments.
1202 1720
1203The code reference is supposed to return a callback (usually a closure) 1721The function is supposed to return a callback (usually a closure) that
1204that works as a plain read callback (see C<< ->push_read ($cb) >>). 1722works as a plain read callback (see C<< ->push_read ($cb) >>), so you can
1723mentally treat the function as a "configurable read type to read callback"
1724converter.
1205 1725
1206It should invoke the passed callback when it is done reading (remember to 1726It should invoke the original callback when it is done reading (remember
1207pass C<$handle> as first argument as all other callbacks do that). 1727to pass C<$handle> as first argument as all other callbacks do that,
1728although there is no strict requirement on this).
1208 1729
1209Note that this is a function, and all types registered this way will be
1210global, so try to use unique names.
1211
1212For examples, see the source of this module (F<perldoc -m AnyEvent::Handle>, 1730For examples, see the source of this module (F<perldoc -m
1213search for C<register_read_type>)). 1731AnyEvent::Handle>, search for C<register_read_type>)).
1214 1732
1215=item $handle->stop_read 1733=item $handle->stop_read
1216 1734
1217=item $handle->start_read 1735=item $handle->start_read
1218 1736
1224Note that AnyEvent::Handle will automatically C<start_read> for you when 1742Note that AnyEvent::Handle will automatically C<start_read> for you when
1225you change the C<on_read> callback or push/unshift a read callback, and it 1743you change the C<on_read> callback or push/unshift a read callback, and it
1226will automatically C<stop_read> for you when neither C<on_read> is set nor 1744will automatically C<stop_read> for you when neither C<on_read> is set nor
1227there are any read requests in the queue. 1745there are any read requests in the queue.
1228 1746
1747These methods will have no effect when in TLS mode (as TLS doesn't support
1748half-duplex connections).
1749
1229=cut 1750=cut
1230 1751
1231sub stop_read { 1752sub stop_read {
1232 my ($self) = @_; 1753 my ($self) = @_;
1233 1754
1234 delete $self->{_rw}; 1755 delete $self->{_rw} unless $self->{tls};
1235} 1756}
1236 1757
1237sub start_read { 1758sub start_read {
1238 my ($self) = @_; 1759 my ($self) = @_;
1239 1760
1240 unless ($self->{_rw} || $self->{_eof}) { 1761 unless ($self->{_rw} || $self->{_eof} || !$self->{fh}) {
1241 Scalar::Util::weaken $self; 1762 Scalar::Util::weaken $self;
1242 1763
1243 $self->{_rw} = AnyEvent->io (fh => $self->{fh}, poll => "r", cb => sub { 1764 $self->{_rw} = AE::io $self->{fh}, 0, sub {
1244 my $rbuf = $self->{filter_r} ? \my $buf : \$self->{rbuf}; 1765 my $rbuf = \($self->{tls} ? my $buf : $self->{rbuf});
1245 my $len = sysread $self->{fh}, $$rbuf, $self->{read_size} || 8192, length $$rbuf; 1766 my $len = sysread $self->{fh}, $$rbuf, $self->{read_size}, length $$rbuf;
1246 1767
1247 if ($len > 0) { 1768 if ($len > 0) {
1248 $self->{_activity} = AnyEvent->now; 1769 $self->{_activity} = $self->{_ractivity} = AE::now;
1249 1770
1250 $self->{filter_r} 1771 if ($self->{tls}) {
1251 ? $self->{filter_r}($self, $rbuf) 1772 Net::SSLeay::BIO_write ($self->{_rbio}, $$rbuf);
1252 : $self->{_in_drain} || $self->_drain_rbuf; 1773
1774 &_dotls ($self);
1775 } else {
1776 $self->_drain_rbuf;
1777 }
1778
1779 if ($len == $self->{read_size}) {
1780 $self->{read_size} *= 2;
1781 $self->{read_size} = $self->{max_read_size} || MAX_READ_SIZE
1782 if $self->{read_size} > ($self->{max_read_size} || MAX_READ_SIZE);
1783 }
1253 1784
1254 } elsif (defined $len) { 1785 } elsif (defined $len) {
1255 delete $self->{_rw}; 1786 delete $self->{_rw};
1256 $self->{_eof} = 1; 1787 $self->{_eof} = 1;
1257 $self->_drain_rbuf unless $self->{_in_drain}; 1788 $self->_drain_rbuf;
1258 1789
1259 } elsif ($! != EAGAIN && $! != EINTR && $! != WSAEWOULDBLOCK) { 1790 } elsif ($! != EAGAIN && $! != EINTR && $! != WSAEWOULDBLOCK) {
1260 return $self->_error ($!, 1); 1791 return $self->_error ($!, 1);
1261 } 1792 }
1262 }); 1793 };
1263 } 1794 }
1264} 1795}
1265 1796
1797our $ERROR_SYSCALL;
1798our $ERROR_WANT_READ;
1799
1800sub _tls_error {
1801 my ($self, $err) = @_;
1802
1803 return $self->_error ($!, 1)
1804 if $err == Net::SSLeay::ERROR_SYSCALL ();
1805
1806 my $err =Net::SSLeay::ERR_error_string (Net::SSLeay::ERR_get_error ());
1807
1808 # reduce error string to look less scary
1809 $err =~ s/^error:[0-9a-fA-F]{8}:[^:]+:([^:]+):/\L$1: /;
1810
1811 if ($self->{_on_starttls}) {
1812 (delete $self->{_on_starttls})->($self, undef, $err);
1813 &_freetls;
1814 } else {
1815 &_freetls;
1816 $self->_error (Errno::EPROTO, 1, $err);
1817 }
1818}
1819
1820# poll the write BIO and send the data if applicable
1821# also decode read data if possible
1822# this is basiclaly our TLS state machine
1823# more efficient implementations are possible with openssl,
1824# but not with the buggy and incomplete Net::SSLeay.
1266sub _dotls { 1825sub _dotls {
1267 my ($self) = @_; 1826 my ($self) = @_;
1268 1827
1269 my $buf; 1828 my $tmp;
1270 1829
1271 if (length $self->{_tls_wbuf}) { 1830 if (length $self->{_tls_wbuf}) {
1272 while ((my $len = Net::SSLeay::write ($self->{tls}, $self->{_tls_wbuf})) > 0) { 1831 while (($tmp = Net::SSLeay::write ($self->{tls}, $self->{_tls_wbuf})) > 0) {
1273 substr $self->{_tls_wbuf}, 0, $len, ""; 1832 substr $self->{_tls_wbuf}, 0, $tmp, "";
1274 } 1833 }
1275 }
1276 1834
1835 $tmp = Net::SSLeay::get_error ($self->{tls}, $tmp);
1836 return $self->_tls_error ($tmp)
1837 if $tmp != $ERROR_WANT_READ
1838 && ($tmp != $ERROR_SYSCALL || $!);
1839 }
1840
1841 while (defined ($tmp = Net::SSLeay::read ($self->{tls}))) {
1842 unless (length $tmp) {
1843 $self->{_on_starttls}
1844 and (delete $self->{_on_starttls})->($self, undef, "EOF during handshake"); # ???
1845 &_freetls;
1846
1847 if ($self->{on_stoptls}) {
1848 $self->{on_stoptls}($self);
1849 return;
1850 } else {
1851 # let's treat SSL-eof as we treat normal EOF
1852 delete $self->{_rw};
1853 $self->{_eof} = 1;
1854 }
1855 }
1856
1857 $self->{_tls_rbuf} .= $tmp;
1858 $self->_drain_rbuf;
1859 $self->{tls} or return; # tls session might have gone away in callback
1860 }
1861
1862 $tmp = Net::SSLeay::get_error ($self->{tls}, -1);
1863 return $self->_tls_error ($tmp)
1864 if $tmp != $ERROR_WANT_READ
1865 && ($tmp != $ERROR_SYSCALL || $!);
1866
1277 if (length ($buf = Net::SSLeay::BIO_read ($self->{_wbio}))) { 1867 while (length ($tmp = Net::SSLeay::BIO_read ($self->{_wbio}))) {
1278 $self->{wbuf} .= $buf; 1868 $self->{wbuf} .= $tmp;
1279 $self->_drain_wbuf; 1869 $self->_drain_wbuf;
1870 $self->{tls} or return; # tls session might have gone away in callback
1280 } 1871 }
1281 1872
1282 while (defined ($buf = Net::SSLeay::read ($self->{tls}))) { 1873 $self->{_on_starttls}
1283 if (length $buf) { 1874 and Net::SSLeay::state ($self->{tls}) == Net::SSLeay::ST_OK ()
1284 $self->{rbuf} .= $buf; 1875 and (delete $self->{_on_starttls})->($self, 1, "TLS/SSL connection established");
1285 $self->_drain_rbuf unless $self->{_in_drain};
1286 } else {
1287 # let's treat SSL-eof as we treat normal EOF
1288 $self->{_eof} = 1;
1289 $self->_shutdown;
1290 return;
1291 }
1292 }
1293
1294 my $err = Net::SSLeay::get_error ($self->{tls}, -1);
1295
1296 if ($err!= Net::SSLeay::ERROR_WANT_READ ()) {
1297 if ($err == Net::SSLeay::ERROR_SYSCALL ()) {
1298 return $self->_error ($!, 1);
1299 } elsif ($err == Net::SSLeay::ERROR_SSL ()) {
1300 return $self->_error (&Errno::EIO, 1);
1301 }
1302
1303 # all others are fine for our purposes
1304 }
1305} 1876}
1306 1877
1307=item $handle->starttls ($tls[, $tls_ctx]) 1878=item $handle->starttls ($tls[, $tls_ctx])
1308 1879
1309Instead of starting TLS negotiation immediately when the AnyEvent::Handle 1880Instead of starting TLS negotiation immediately when the AnyEvent::Handle
1310object is created, you can also do that at a later time by calling 1881object is created, you can also do that at a later time by calling
1311C<starttls>. 1882C<starttls>.
1312 1883
1884Starting TLS is currently an asynchronous operation - when you push some
1885write data and then call C<< ->starttls >> then TLS negotiation will start
1886immediately, after which the queued write data is then sent.
1887
1313The first argument is the same as the C<tls> constructor argument (either 1888The first argument is the same as the C<tls> constructor argument (either
1314C<"connect">, C<"accept"> or an existing Net::SSLeay object). 1889C<"connect">, C<"accept"> or an existing Net::SSLeay object).
1315 1890
1316The second argument is the optional C<Net::SSLeay::CTX> object that is 1891The second argument is the optional C<AnyEvent::TLS> object that is used
1317used when AnyEvent::Handle has to create its own TLS connection object. 1892when AnyEvent::Handle has to create its own TLS connection object, or
1893a hash reference with C<< key => value >> pairs that will be used to
1894construct a new context.
1318 1895
1319The TLS connection object will end up in C<< $handle->{tls} >> after this 1896The TLS connection object will end up in C<< $handle->{tls} >>, the TLS
1320call and can be used or changed to your liking. Note that the handshake 1897context in C<< $handle->{tls_ctx} >> after this call and can be used or
1321might have already started when this function returns. 1898changed to your liking. Note that the handshake might have already started
1899when this function returns.
1322 1900
1901Due to bugs in OpenSSL, it might or might not be possible to do multiple
1902handshakes on the same stream. It is best to not attempt to use the
1903stream after stopping TLS.
1904
1905This method may invoke callbacks (and therefore the handle might be
1906destroyed after it returns).
1907
1323=cut 1908=cut
1909
1910our %TLS_CACHE; #TODO not yet documented, should we?
1324 1911
1325sub starttls { 1912sub starttls {
1326 my ($self, $ssl, $ctx) = @_; 1913 my ($self, $tls, $ctx) = @_;
1327 1914
1328 $self->stoptls; 1915 Carp::croak "It is an error to call starttls on an AnyEvent::Handle object while TLS is already active, caught"
1916 if $self->{tls};
1329 1917
1330 if ($ssl eq "accept") { 1918 $self->{tls} = $tls;
1331 $ssl = Net::SSLeay::new ($ctx || TLS_CTX ()); 1919 $self->{tls_ctx} = $ctx if @_ > 2;
1332 Net::SSLeay::set_accept_state ($ssl); 1920
1333 } elsif ($ssl eq "connect") { 1921 return unless $self->{fh};
1334 $ssl = Net::SSLeay::new ($ctx || TLS_CTX ()); 1922
1335 Net::SSLeay::set_connect_state ($ssl); 1923 require Net::SSLeay;
1924
1925 $ERROR_SYSCALL = Net::SSLeay::ERROR_SYSCALL ();
1926 $ERROR_WANT_READ = Net::SSLeay::ERROR_WANT_READ ();
1927
1928 $tls = delete $self->{tls};
1929 $ctx = $self->{tls_ctx};
1930
1931 local $Carp::CarpLevel = 1; # skip ourselves when creating a new context or session
1932
1933 if ("HASH" eq ref $ctx) {
1934 require AnyEvent::TLS;
1935
1936 if ($ctx->{cache}) {
1937 my $key = $ctx+0;
1938 $ctx = $TLS_CACHE{$key} ||= new AnyEvent::TLS %$ctx;
1939 } else {
1940 $ctx = new AnyEvent::TLS %$ctx;
1941 }
1942 }
1336 } 1943
1337 1944 $self->{tls_ctx} = $ctx || TLS_CTX ();
1338 $self->{tls} = $ssl; 1945 $self->{tls} = $tls = $self->{tls_ctx}->_get_session ($tls, $self, $self->{peername});
1339 1946
1340 # basically, this is deep magic (because SSL_read should have the same issues) 1947 # basically, this is deep magic (because SSL_read should have the same issues)
1341 # but the openssl maintainers basically said: "trust us, it just works". 1948 # but the openssl maintainers basically said: "trust us, it just works".
1342 # (unfortunately, we have to hardcode constants because the abysmally misdesigned 1949 # (unfortunately, we have to hardcode constants because the abysmally misdesigned
1343 # and mismaintained ssleay-module doesn't even offer them). 1950 # and mismaintained ssleay-module doesn't even offer them).
1344 # http://www.mail-archive.com/openssl-dev@openssl.org/msg22420.html 1951 # http://www.mail-archive.com/openssl-dev@openssl.org/msg22420.html
1952 #
1953 # in short: this is a mess.
1954 #
1955 # note that we do not try to keep the length constant between writes as we are required to do.
1956 # we assume that most (but not all) of this insanity only applies to non-blocking cases,
1957 # and we drive openssl fully in blocking mode here. Or maybe we don't - openssl seems to
1958 # have identity issues in that area.
1345 Net::SSLeay::CTX_set_mode ($self->{tls}, 1959# Net::SSLeay::CTX_set_mode ($ssl,
1346 (eval { local $SIG{__DIE__}; Net::SSLeay::MODE_ENABLE_PARTIAL_WRITE () } || 1) 1960# (eval { local $SIG{__DIE__}; Net::SSLeay::MODE_ENABLE_PARTIAL_WRITE () } || 1)
1347 | (eval { local $SIG{__DIE__}; Net::SSLeay::MODE_ACCEPT_MOVING_WRITE_BUFFER () } || 2)); 1961# | (eval { local $SIG{__DIE__}; Net::SSLeay::MODE_ACCEPT_MOVING_WRITE_BUFFER () } || 2));
1962 Net::SSLeay::CTX_set_mode ($tls, 1|2);
1348 1963
1349 $self->{_rbio} = Net::SSLeay::BIO_new (Net::SSLeay::BIO_s_mem ()); 1964 $self->{_rbio} = Net::SSLeay::BIO_new (Net::SSLeay::BIO_s_mem ());
1350 $self->{_wbio} = Net::SSLeay::BIO_new (Net::SSLeay::BIO_s_mem ()); 1965 $self->{_wbio} = Net::SSLeay::BIO_new (Net::SSLeay::BIO_s_mem ());
1351 1966
1967 Net::SSLeay::BIO_write ($self->{_rbio}, delete $self->{rbuf});
1968
1352 Net::SSLeay::set_bio ($ssl, $self->{_rbio}, $self->{_wbio}); 1969 Net::SSLeay::set_bio ($tls, $self->{_rbio}, $self->{_wbio});
1353 1970
1354 $self->{filter_w} = sub { 1971 $self->{_on_starttls} = sub { $_[0]{on_starttls}(@_) }
1355 $_[0]{_tls_wbuf} .= ${$_[1]}; 1972 if $self->{on_starttls};
1356 &_dotls; 1973
1357 }; 1974 &_dotls; # need to trigger the initial handshake
1358 $self->{filter_r} = sub { 1975 $self->start_read; # make sure we actually do read
1359 Net::SSLeay::BIO_write ($_[0]{_rbio}, ${$_[1]});
1360 &_dotls;
1361 };
1362} 1976}
1363 1977
1364=item $handle->stoptls 1978=item $handle->stoptls
1365 1979
1366Destroys the SSL connection, if any. Partial read or write data will be 1980Shuts down the SSL connection - this makes a proper EOF handshake by
1367lost. 1981sending a close notify to the other side, but since OpenSSL doesn't
1982support non-blocking shut downs, it is not guaranteed that you can re-use
1983the stream afterwards.
1984
1985This method may invoke callbacks (and therefore the handle might be
1986destroyed after it returns).
1368 1987
1369=cut 1988=cut
1370 1989
1371sub stoptls { 1990sub stoptls {
1372 my ($self) = @_; 1991 my ($self) = @_;
1373 1992
1374 Net::SSLeay::free (delete $self->{tls}) if $self->{tls}; 1993 if ($self->{tls} && $self->{fh}) {
1994 Net::SSLeay::shutdown ($self->{tls});
1375 1995
1376 delete $self->{_rbio}; 1996 &_dotls;
1377 delete $self->{_wbio}; 1997
1378 delete $self->{_tls_wbuf}; 1998# # we don't give a shit. no, we do, but we can't. no...#d#
1379 delete $self->{filter_r}; 1999# # we, we... have to use openssl :/#d#
1380 delete $self->{filter_w}; 2000# &_freetls;#d#
2001 }
2002}
2003
2004sub _freetls {
2005 my ($self) = @_;
2006
2007 return unless $self->{tls};
2008
2009 $self->{tls_ctx}->_put_session (delete $self->{tls})
2010 if $self->{tls} > 0;
2011
2012 delete @$self{qw(_rbio _wbio _tls_wbuf _on_starttls)};
1381} 2013}
1382 2014
1383sub DESTROY { 2015sub DESTROY {
1384 my $self = shift; 2016 my ($self) = @_;
1385 2017
1386 $self->stoptls; 2018 &_freetls;
1387 2019
1388 my $linger = exists $self->{linger} ? $self->{linger} : 3600; 2020 my $linger = exists $self->{linger} ? $self->{linger} : 3600;
1389 2021
1390 if ($linger && length $self->{wbuf}) { 2022 if ($linger && length $self->{wbuf} && $self->{fh}) {
1391 my $fh = delete $self->{fh}; 2023 my $fh = delete $self->{fh};
1392 my $wbuf = delete $self->{wbuf}; 2024 my $wbuf = delete $self->{wbuf};
1393 2025
1394 my @linger; 2026 my @linger;
1395 2027
1396 push @linger, AnyEvent->io (fh => $fh, poll => "w", cb => sub { 2028 push @linger, AE::io $fh, 1, sub {
1397 my $len = syswrite $fh, $wbuf, length $wbuf; 2029 my $len = syswrite $fh, $wbuf, length $wbuf;
1398 2030
1399 if ($len > 0) { 2031 if ($len > 0) {
1400 substr $wbuf, 0, $len, ""; 2032 substr $wbuf, 0, $len, "";
1401 } else { 2033 } elsif (defined $len || ($! != EAGAIN && $! != EINTR && $! != WSAEWOULDBLOCK)) {
1402 @linger = (); # end 2034 @linger = (); # end
1403 } 2035 }
1404 }); 2036 };
1405 push @linger, AnyEvent->timer (after => $linger, cb => sub { 2037 push @linger, AE::timer $linger, 0, sub {
1406 @linger = (); 2038 @linger = ();
1407 }); 2039 };
1408 } 2040 }
1409} 2041}
2042
2043=item $handle->destroy
2044
2045Shuts down the handle object as much as possible - this call ensures that
2046no further callbacks will be invoked and as many resources as possible
2047will be freed. Any method you will call on the handle object after
2048destroying it in this way will be silently ignored (and it will return the
2049empty list).
2050
2051Normally, you can just "forget" any references to an AnyEvent::Handle
2052object and it will simply shut down. This works in fatal error and EOF
2053callbacks, as well as code outside. It does I<NOT> work in a read or write
2054callback, so when you want to destroy the AnyEvent::Handle object from
2055within such an callback. You I<MUST> call C<< ->destroy >> explicitly in
2056that case.
2057
2058Destroying the handle object in this way has the advantage that callbacks
2059will be removed as well, so if those are the only reference holders (as
2060is common), then one doesn't need to do anything special to break any
2061reference cycles.
2062
2063The handle might still linger in the background and write out remaining
2064data, as specified by the C<linger> option, however.
2065
2066=cut
2067
2068sub destroy {
2069 my ($self) = @_;
2070
2071 $self->DESTROY;
2072 %$self = ();
2073 bless $self, "AnyEvent::Handle::destroyed";
2074}
2075
2076sub AnyEvent::Handle::destroyed::AUTOLOAD {
2077 #nop
2078}
2079
2080=item $handle->destroyed
2081
2082Returns false as long as the handle hasn't been destroyed by a call to C<<
2083->destroy >>, true otherwise.
2084
2085Can be useful to decide whether the handle is still valid after some
2086callback possibly destroyed the handle. For example, C<< ->push_write >>,
2087C<< ->starttls >> and other methods can call user callbacks, which in turn
2088can destroy the handle, so work can be avoided by checking sometimes:
2089
2090 $hdl->starttls ("accept");
2091 return if $hdl->destroyed;
2092 $hdl->push_write (...
2093
2094Note that the call to C<push_write> will silently be ignored if the handle
2095has been destroyed, so often you can just ignore the possibility of the
2096handle being destroyed.
2097
2098=cut
2099
2100sub destroyed { 0 }
2101sub AnyEvent::Handle::destroyed::destroyed { 1 }
1410 2102
1411=item AnyEvent::Handle::TLS_CTX 2103=item AnyEvent::Handle::TLS_CTX
1412 2104
1413This function creates and returns the Net::SSLeay::CTX object used by 2105This function creates and returns the AnyEvent::TLS object used by default
1414default for TLS mode. 2106for TLS mode.
1415 2107
1416The context is created like this: 2108The context is created by calling L<AnyEvent::TLS> without any arguments.
1417
1418 Net::SSLeay::load_error_strings;
1419 Net::SSLeay::SSLeay_add_ssl_algorithms;
1420 Net::SSLeay::randomize;
1421
1422 my $CTX = Net::SSLeay::CTX_new;
1423
1424 Net::SSLeay::CTX_set_options $CTX, Net::SSLeay::OP_ALL
1425 2109
1426=cut 2110=cut
1427 2111
1428our $TLS_CTX; 2112our $TLS_CTX;
1429 2113
1430sub TLS_CTX() { 2114sub TLS_CTX() {
1431 $TLS_CTX || do { 2115 $TLS_CTX ||= do {
1432 require Net::SSLeay; 2116 require AnyEvent::TLS;
1433 2117
1434 Net::SSLeay::load_error_strings (); 2118 new AnyEvent::TLS
1435 Net::SSLeay::SSLeay_add_ssl_algorithms ();
1436 Net::SSLeay::randomize ();
1437
1438 $TLS_CTX = Net::SSLeay::CTX_new ();
1439
1440 Net::SSLeay::CTX_set_options ($TLS_CTX, Net::SSLeay::OP_ALL ());
1441
1442 $TLS_CTX
1443 } 2119 }
1444} 2120}
1445 2121
1446=back 2122=back
2123
2124
2125=head1 NONFREQUENTLY ASKED QUESTIONS
2126
2127=over 4
2128
2129=item I C<undef> the AnyEvent::Handle reference inside my callback and
2130still get further invocations!
2131
2132That's because AnyEvent::Handle keeps a reference to itself when handling
2133read or write callbacks.
2134
2135It is only safe to "forget" the reference inside EOF or error callbacks,
2136from within all other callbacks, you need to explicitly call the C<<
2137->destroy >> method.
2138
2139=item I get different callback invocations in TLS mode/Why can't I pause
2140reading?
2141
2142Unlike, say, TCP, TLS connections do not consist of two independent
2143communication channels, one for each direction. Or put differently, the
2144read and write directions are not independent of each other: you cannot
2145write data unless you are also prepared to read, and vice versa.
2146
2147This means that, in TLS mode, you might get C<on_error> or C<on_eof>
2148callback invocations when you are not expecting any read data - the reason
2149is that AnyEvent::Handle always reads in TLS mode.
2150
2151During the connection, you have to make sure that you always have a
2152non-empty read-queue, or an C<on_read> watcher. At the end of the
2153connection (or when you no longer want to use it) you can call the
2154C<destroy> method.
2155
2156=item How do I read data until the other side closes the connection?
2157
2158If you just want to read your data into a perl scalar, the easiest way
2159to achieve this is by setting an C<on_read> callback that does nothing,
2160clearing the C<on_eof> callback and in the C<on_error> callback, the data
2161will be in C<$_[0]{rbuf}>:
2162
2163 $handle->on_read (sub { });
2164 $handle->on_eof (undef);
2165 $handle->on_error (sub {
2166 my $data = delete $_[0]{rbuf};
2167 });
2168
2169The reason to use C<on_error> is that TCP connections, due to latencies
2170and packets loss, might get closed quite violently with an error, when in
2171fact all data has been received.
2172
2173It is usually better to use acknowledgements when transferring data,
2174to make sure the other side hasn't just died and you got the data
2175intact. This is also one reason why so many internet protocols have an
2176explicit QUIT command.
2177
2178=item I don't want to destroy the handle too early - how do I wait until
2179all data has been written?
2180
2181After writing your last bits of data, set the C<on_drain> callback
2182and destroy the handle in there - with the default setting of
2183C<low_water_mark> this will be called precisely when all data has been
2184written to the socket:
2185
2186 $handle->push_write (...);
2187 $handle->on_drain (sub {
2188 warn "all data submitted to the kernel\n";
2189 undef $handle;
2190 });
2191
2192If you just want to queue some data and then signal EOF to the other side,
2193consider using C<< ->push_shutdown >> instead.
2194
2195=item I want to contact a TLS/SSL server, I don't care about security.
2196
2197If your TLS server is a pure TLS server (e.g. HTTPS) that only speaks TLS,
2198connect to it and then create the AnyEvent::Handle with the C<tls>
2199parameter:
2200
2201 tcp_connect $host, $port, sub {
2202 my ($fh) = @_;
2203
2204 my $handle = new AnyEvent::Handle
2205 fh => $fh,
2206 tls => "connect",
2207 on_error => sub { ... };
2208
2209 $handle->push_write (...);
2210 };
2211
2212=item I want to contact a TLS/SSL server, I do care about security.
2213
2214Then you should additionally enable certificate verification, including
2215peername verification, if the protocol you use supports it (see
2216L<AnyEvent::TLS>, C<verify_peername>).
2217
2218E.g. for HTTPS:
2219
2220 tcp_connect $host, $port, sub {
2221 my ($fh) = @_;
2222
2223 my $handle = new AnyEvent::Handle
2224 fh => $fh,
2225 peername => $host,
2226 tls => "connect",
2227 tls_ctx => { verify => 1, verify_peername => "https" },
2228 ...
2229
2230Note that you must specify the hostname you connected to (or whatever
2231"peername" the protocol needs) as the C<peername> argument, otherwise no
2232peername verification will be done.
2233
2234The above will use the system-dependent default set of trusted CA
2235certificates. If you want to check against a specific CA, add the
2236C<ca_file> (or C<ca_cert>) arguments to C<tls_ctx>:
2237
2238 tls_ctx => {
2239 verify => 1,
2240 verify_peername => "https",
2241 ca_file => "my-ca-cert.pem",
2242 },
2243
2244=item I want to create a TLS/SSL server, how do I do that?
2245
2246Well, you first need to get a server certificate and key. You have
2247three options: a) ask a CA (buy one, use cacert.org etc.) b) create a
2248self-signed certificate (cheap. check the search engine of your choice,
2249there are many tutorials on the net) or c) make your own CA (tinyca2 is a
2250nice program for that purpose).
2251
2252Then create a file with your private key (in PEM format, see
2253L<AnyEvent::TLS>), followed by the certificate (also in PEM format). The
2254file should then look like this:
2255
2256 -----BEGIN RSA PRIVATE KEY-----
2257 ...header data
2258 ... lots of base64'y-stuff
2259 -----END RSA PRIVATE KEY-----
2260
2261 -----BEGIN CERTIFICATE-----
2262 ... lots of base64'y-stuff
2263 -----END CERTIFICATE-----
2264
2265The important bits are the "PRIVATE KEY" and "CERTIFICATE" parts. Then
2266specify this file as C<cert_file>:
2267
2268 tcp_server undef, $port, sub {
2269 my ($fh) = @_;
2270
2271 my $handle = new AnyEvent::Handle
2272 fh => $fh,
2273 tls => "accept",
2274 tls_ctx => { cert_file => "my-server-keycert.pem" },
2275 ...
2276
2277When you have intermediate CA certificates that your clients might not
2278know about, just append them to the C<cert_file>.
2279
2280=back
2281
1447 2282
1448=head1 SUBCLASSING AnyEvent::Handle 2283=head1 SUBCLASSING AnyEvent::Handle
1449 2284
1450In many cases, you might want to subclass AnyEvent::Handle. 2285In many cases, you might want to subclass AnyEvent::Handle.
1451 2286
1468 2303
1469=item * all members not documented here and not prefixed with an underscore 2304=item * all members not documented here and not prefixed with an underscore
1470are free to use in subclasses. 2305are free to use in subclasses.
1471 2306
1472Of course, new versions of AnyEvent::Handle may introduce more "public" 2307Of course, new versions of AnyEvent::Handle may introduce more "public"
1473member variables, but thats just life, at least it is documented. 2308member variables, but that's just life. At least it is documented.
1474 2309
1475=back 2310=back
1476 2311
1477=head1 AUTHOR 2312=head1 AUTHOR
1478 2313

Diff Legend

Removed lines
+ Added lines
< Changed lines
> Changed lines