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 1.8 by root, Fri May 2 15:36:10 2008 UTC vs.
Revision 1.58 by root, Wed Jun 4 22:51:15 2008 UTC

2 2
3no warnings; 3no warnings;
4use strict; 4use strict;
5 5
6use AnyEvent (); 6use AnyEvent ();
7use AnyEvent::Util (); 7use AnyEvent::Util qw(WSAEWOULDBLOCK);
8use Scalar::Util (); 8use Scalar::Util ();
9use Carp (); 9use Carp ();
10use Fcntl (); 10use Fcntl ();
11use Errno qw/EAGAIN EINTR/; 11use Errno qw(EAGAIN EINTR);
12 12
13=head1 NAME 13=head1 NAME
14 14
15AnyEvent::Handle - non-blocking I/O on filehandles via AnyEvent 15AnyEvent::Handle - non-blocking I/O on file handles via AnyEvent
16 16
17=cut 17=cut
18 18
19our $VERSION = '0.02'; 19our $VERSION = 4.13;
20 20
21=head1 SYNOPSIS 21=head1 SYNOPSIS
22 22
23 use AnyEvent; 23 use AnyEvent;
24 use AnyEvent::Handle; 24 use AnyEvent::Handle;
25 25
26 my $cv = AnyEvent->condvar; 26 my $cv = AnyEvent->condvar;
27 27
28 my $ae_fh = AnyEvent::Handle->new (fh => \*STDIN); 28 my $handle =
29
30 #TODO
31
32 # or use the constructor to pass the callback:
33
34 my $ae_fh2 =
35 AnyEvent::Handle->new ( 29 AnyEvent::Handle->new (
36 fh => \*STDIN, 30 fh => \*STDIN,
37 on_eof => sub { 31 on_eof => sub {
38 $cv->broadcast; 32 $cv->broadcast;
39 }, 33 },
40 #TODO
41 ); 34 );
42 35
43 $cv->wait; 36 # send some request line
37 $handle->push_write ("getinfo\015\012");
38
39 # read the response line
40 $handle->push_read (line => sub {
41 my ($handle, $line) = @_;
42 warn "read line <$line>\n";
43 $cv->send;
44 });
45
46 $cv->recv;
44 47
45=head1 DESCRIPTION 48=head1 DESCRIPTION
46 49
47This module is a helper module to make it easier to do event-based I/O on 50This module is a helper module to make it easier to do event-based I/O on
48filehandles (and sockets, see L<AnyEvent::Socket> for an easy way to make 51filehandles. For utility functions for doing non-blocking connects and accepts
49non-blocking resolves and connects). 52on sockets see L<AnyEvent::Util>.
50 53
51In the following, when the documentation refers to of "bytes" then this 54In the following, when the documentation refers to of "bytes" then this
52means characters. As sysread and syswrite are used for all I/O, their 55means characters. As sysread and syswrite are used for all I/O, their
53treatment of characters applies to this module as well. 56treatment of characters applies to this module as well.
54 57
70The filehandle this L<AnyEvent::Handle> object will operate on. 73The filehandle this L<AnyEvent::Handle> object will operate on.
71 74
72NOTE: The filehandle will be set to non-blocking (using 75NOTE: The filehandle will be set to non-blocking (using
73AnyEvent::Util::fh_nonblocking). 76AnyEvent::Util::fh_nonblocking).
74 77
75=item on_error => $cb->($self) [MANDATORY] 78=item on_eof => $cb->($handle)
76 79
77This is the fatal error callback, that is called when a fatal error ocurs, 80Set the callback to be called when an end-of-file condition is detcted,
81i.e. in the case of a socket, when the other side has closed the
82connection cleanly.
83
84While not mandatory, it is highly recommended to set an eof callback,
85otherwise you might end up with a closed socket while you are still
86waiting for data.
87
88=item on_error => $cb->($handle, $fatal)
89
90This is the error callback, which is called when, well, some error
78such as not being able to resolve the hostname, failure to connect or a 91occured, such as not being able to resolve the hostname, failure to
79read error. 92connect or a read error.
80 93
81The object will not be in a usable state when this callback has been 94Some errors are fatal (which is indicated by C<$fatal> being true). On
82called. 95fatal errors the handle object will be shut down and will not be
96usable. Non-fatal errors can be retried by simply returning, but it is
97recommended to simply ignore this parameter and instead abondon the handle
98object when this callback is invoked.
83 99
84On callback entrance, the value of C<$!> contains the opertaing system 100On callback entrance, the value of C<$!> contains the operating system
85error (or C<ENOSPC> or C<EPIPE>). 101error (or C<ENOSPC>, C<EPIPE>, C<ETIMEDOUT> or C<EBADMSG>).
86 102
87=item on_eof => $cb->($self) [MANDATORY] 103While not mandatory, it is I<highly> recommended to set this callback, as
104you will not be notified of errors otherwise. The default simply calls
105C<croak>.
88 106
89Set the callback to be called on EOF.
90
91=item on_read => $cb->($self) 107=item on_read => $cb->($handle)
92 108
93This sets the default read callback, which is called when data arrives 109This sets the default read callback, which is called when data arrives
94and no read request is in the queue. If the read callback is C<undef> 110and no read request is in the queue.
95or has never been set, than AnyEvent::Handle will cease reading from the
96filehandle.
97 111
98To access (and remove data from) the read buffer, use the C<< ->rbuf >> 112To access (and remove data from) the read buffer, use the C<< ->rbuf >>
99method or acces sthe C<$self->{rbuf}> member directly. 113method or access the C<$handle->{rbuf}> member directly.
100 114
101When an EOF condition is detected then AnyEvent::Handle will first try to 115When an EOF condition is detected then AnyEvent::Handle will first try to
102feed all the remaining data to the queued callbacks and C<on_read> before 116feed all the remaining data to the queued callbacks and C<on_read> before
103calling the C<on_eof> callback. If no progress can be made, then a fatal 117calling the C<on_eof> callback. If no progress can be made, then a fatal
104error will be raised (with C<$!> set to C<EPIPE>). 118error will be raised (with C<$!> set to C<EPIPE>).
105 119
106=item on_drain => $cb->() 120=item on_drain => $cb->($handle)
107 121
108This sets the callback that is called when the write buffer becomes empty 122This sets the callback that is called when the write buffer becomes empty
109(or when the callback is set and the buffer is empty already). 123(or when the callback is set and the buffer is empty already).
110 124
111To append to the write buffer, use the C<< ->push_write >> method. 125To append to the write buffer, use the C<< ->push_write >> method.
126
127=item timeout => $fractional_seconds
128
129If non-zero, then this enables an "inactivity" timeout: whenever this many
130seconds pass without a successful read or write on the underlying file
131handle, the C<on_timeout> callback will be invoked (and if that one is
132missing, an C<ETIMEDOUT> error will be raised).
133
134Note that timeout processing is also active when you currently do not have
135any outstanding read or write requests: If you plan to keep the connection
136idle then you should disable the timout temporarily or ignore the timeout
137in the C<on_timeout> callback.
138
139Zero (the default) disables this timeout.
140
141=item on_timeout => $cb->($handle)
142
143Called whenever the inactivity timeout passes. If you return from this
144callback, then the timeout will be reset as if some activity had happened,
145so this condition is not fatal in any way.
112 146
113=item rbuf_max => <bytes> 147=item rbuf_max => <bytes>
114 148
115If defined, then a fatal error will be raised (with C<$!> set to C<ENOSPC>) 149If defined, then a fatal error will be raised (with C<$!> set to C<ENOSPC>)
116when the read buffer ever (strictly) exceeds this size. This is useful to 150when the read buffer ever (strictly) exceeds this size. This is useful to
123isn't finished). 157isn't finished).
124 158
125=item read_size => <bytes> 159=item read_size => <bytes>
126 160
127The default read block size (the amount of bytes this module will try to read 161The default read block size (the amount of bytes this module will try to read
128on each [loop iteration). Default: C<4096>. 162during each (loop iteration). Default: C<8192>.
129 163
130=item low_water_mark => <bytes> 164=item low_water_mark => <bytes>
131 165
132Sets the amount of bytes (default: C<0>) that make up an "empty" write 166Sets the amount of bytes (default: C<0>) that make up an "empty" write
133buffer: If the write reaches this size or gets even samller it is 167buffer: If the write reaches this size or gets even samller it is
134considered empty. 168considered empty.
135 169
170=item tls => "accept" | "connect" | Net::SSLeay::SSL object
171
172When this parameter is given, it enables TLS (SSL) mode, that means it
173will start making tls handshake and will transparently encrypt/decrypt
174data.
175
176TLS mode requires Net::SSLeay to be installed (it will be loaded
177automatically when you try to create a TLS handle).
178
179For the TLS server side, use C<accept>, and for the TLS client side of a
180connection, use C<connect> mode.
181
182You can also provide your own TLS connection object, but you have
183to make sure that you call either C<Net::SSLeay::set_connect_state>
184or C<Net::SSLeay::set_accept_state> on it before you pass it to
185AnyEvent::Handle.
186
187See the C<starttls> method if you need to start TLs negotiation later.
188
189=item tls_ctx => $ssl_ctx
190
191Use the given Net::SSLeay::CTX object to create the new TLS connection
192(unless a connection object was specified directly). If this parameter is
193missing, then AnyEvent::Handle will use C<AnyEvent::Handle::TLS_CTX>.
194
195=item json => JSON or JSON::XS object
196
197This is the json coder object used by the C<json> read and write types.
198
199If you don't supply it, then AnyEvent::Handle will create and use a
200suitable one, which will write and expect UTF-8 encoded JSON texts.
201
202Note that you are responsible to depend on the JSON module if you want to
203use this functionality, as AnyEvent does not have a dependency itself.
204
205=item filter_r => $cb
206
207=item filter_w => $cb
208
209These exist, but are undocumented at this time.
210
136=back 211=back
137 212
138=cut 213=cut
139 214
140sub new { 215sub new {
144 219
145 $self->{fh} or Carp::croak "mandatory argument fh is missing"; 220 $self->{fh} or Carp::croak "mandatory argument fh is missing";
146 221
147 AnyEvent::Util::fh_nonblocking $self->{fh}, 1; 222 AnyEvent::Util::fh_nonblocking $self->{fh}, 1;
148 223
149 $self->on_error ((delete $self->{on_error}) or Carp::croak "mandatory argument on_error is missing"); 224 if ($self->{tls}) {
150 $self->on_eof ((delete $self->{on_eof} ) or Carp::croak "mandatory argument on_eof is missing"); 225 require Net::SSLeay;
226 $self->starttls (delete $self->{tls}, delete $self->{tls_ctx});
227 }
228
229 $self->{_activity} = AnyEvent->now;
230 $self->_timeout;
151 231
152 $self->on_drain (delete $self->{on_drain}) if $self->{on_drain}; 232 $self->on_drain (delete $self->{on_drain}) if $self->{on_drain};
153 $self->on_read (delete $self->{on_read} ) if $self->{on_read}; 233 $self->on_read (delete $self->{on_read} ) if $self->{on_read};
154 234
155 $self 235 $self
156} 236}
157 237
158sub _shutdown { 238sub _shutdown {
159 my ($self) = @_; 239 my ($self) = @_;
160 240
241 delete $self->{_tw};
161 delete $self->{rw}; 242 delete $self->{_rw};
162 delete $self->{ww}; 243 delete $self->{_ww};
163 delete $self->{fh}; 244 delete $self->{fh};
164}
165 245
246 $self->stoptls;
247}
248
166sub error { 249sub _error {
167 my ($self) = @_; 250 my ($self, $errno, $fatal) = @_;
168 251
169 {
170 local $!;
171 $self->_shutdown; 252 $self->_shutdown
172 } 253 if $fatal;
173 254
255 $! = $errno;
256
257 if ($self->{on_error}) {
174 $self->{on_error}($self); 258 $self->{on_error}($self, $fatal);
259 } else {
260 Carp::croak "AnyEvent::Handle uncaught error: $!";
261 }
175} 262}
176 263
177=item $fh = $handle->fh 264=item $fh = $handle->fh
178 265
179This method returns the filehandle of the L<AnyEvent::Handle> object. 266This method returns the file handle of the L<AnyEvent::Handle> object.
180 267
181=cut 268=cut
182 269
183sub fh { $_[0]->{fh} } 270sub fh { $_[0]{fh} }
184 271
185=item $handle->on_error ($cb) 272=item $handle->on_error ($cb)
186 273
187Replace the current C<on_error> callback (see the C<on_error> constructor argument). 274Replace the current C<on_error> callback (see the C<on_error> constructor argument).
188 275
196 283
197Replace the current C<on_eof> callback (see the C<on_eof> constructor argument). 284Replace the current C<on_eof> callback (see the C<on_eof> constructor argument).
198 285
199=cut 286=cut
200 287
201#############################################################################
202
203sub on_eof { 288sub on_eof {
204 $_[0]{on_eof} = $_[1]; 289 $_[0]{on_eof} = $_[1];
205} 290}
291
292=item $handle->on_timeout ($cb)
293
294Replace the current C<on_timeout> callback, or disables the callback
295(but not the timeout) if C<$cb> = C<undef>. See C<timeout> constructor
296argument.
297
298=cut
299
300sub on_timeout {
301 $_[0]{on_timeout} = $_[1];
302}
303
304#############################################################################
305
306=item $handle->timeout ($seconds)
307
308Configures (or disables) the inactivity timeout.
309
310=cut
311
312sub timeout {
313 my ($self, $timeout) = @_;
314
315 $self->{timeout} = $timeout;
316 $self->_timeout;
317}
318
319# reset the timeout watcher, as neccessary
320# also check for time-outs
321sub _timeout {
322 my ($self) = @_;
323
324 if ($self->{timeout}) {
325 my $NOW = AnyEvent->now;
326
327 # when would the timeout trigger?
328 my $after = $self->{_activity} + $self->{timeout} - $NOW;
329
330 # now or in the past already?
331 if ($after <= 0) {
332 $self->{_activity} = $NOW;
333
334 if ($self->{on_timeout}) {
335 $self->{on_timeout}($self);
336 } else {
337 $self->_error (&Errno::ETIMEDOUT);
338 }
339
340 # callback could have changed timeout value, optimise
341 return unless $self->{timeout};
342
343 # calculate new after
344 $after = $self->{timeout};
345 }
346
347 Scalar::Util::weaken $self;
348 return unless $self; # ->error could have destroyed $self
349
350 $self->{_tw} ||= AnyEvent->timer (after => $after, cb => sub {
351 delete $self->{_tw};
352 $self->_timeout;
353 });
354 } else {
355 delete $self->{_tw};
356 }
357}
358
359#############################################################################
360
361=back
362
363=head2 WRITE QUEUE
364
365AnyEvent::Handle manages two queues per handle, one for writing and one
366for reading.
367
368The write queue is very simple: you can add data to its end, and
369AnyEvent::Handle will automatically try to get rid of it for you.
370
371When data could be written and the write buffer is shorter then the low
372water mark, the C<on_drain> callback will be invoked.
373
374=over 4
206 375
207=item $handle->on_drain ($cb) 376=item $handle->on_drain ($cb)
208 377
209Sets the C<on_drain> callback or clears it (see the description of 378Sets the C<on_drain> callback or clears it (see the description of
210C<on_drain> in the constructor). 379C<on_drain> in the constructor).
226want (only limited by the available memory), as C<AnyEvent::Handle> 395want (only limited by the available memory), as C<AnyEvent::Handle>
227buffers it independently of the kernel. 396buffers it independently of the kernel.
228 397
229=cut 398=cut
230 399
231sub push_write { 400sub _drain_wbuf {
232 my ($self, $data) = @_; 401 my ($self) = @_;
233 402
234 $self->{wbuf} .= $data; 403 if (!$self->{_ww} && length $self->{wbuf}) {
235 404
236 unless ($self->{ww}) {
237 Scalar::Util::weaken $self; 405 Scalar::Util::weaken $self;
406
238 my $cb = sub { 407 my $cb = sub {
239 my $len = syswrite $self->{fh}, $self->{wbuf}; 408 my $len = syswrite $self->{fh}, $self->{wbuf};
240 409
241 if ($len > 0) { 410 if ($len >= 0) {
242 substr $self->{wbuf}, 0, $len, ""; 411 substr $self->{wbuf}, 0, $len, "";
243 412
413 $self->{_activity} = AnyEvent->now;
244 414
245 $self->{on_drain}($self) 415 $self->{on_drain}($self)
246 if $self->{low_water_mark} >= length $self->{wbuf} 416 if $self->{low_water_mark} >= length $self->{wbuf}
247 && $self->{on_drain}; 417 && $self->{on_drain};
248 418
249 delete $self->{ww} unless length $self->{wbuf}; 419 delete $self->{_ww} unless length $self->{wbuf};
250 } elsif ($! != EAGAIN && $! != EINTR) { 420 } elsif ($! != EAGAIN && $! != EINTR && $! != WSAEWOULDBLOCK) {
251 $self->error; 421 $self->_error ($!, 1);
252 } 422 }
253 }; 423 };
254 424
425 # try to write data immediately
426 $cb->();
427
428 # if still data left in wbuf, we need to poll
255 $self->{ww} = AnyEvent->io (fh => $self->{fh}, poll => "w", cb => $cb); 429 $self->{_ww} = AnyEvent->io (fh => $self->{fh}, poll => "w", cb => $cb)
256 430 if length $self->{wbuf};
257 $cb->($self);
258 }; 431 };
259} 432}
260 433
434our %WH;
435
436sub register_write_type($$) {
437 $WH{$_[0]} = $_[1];
438}
439
440sub push_write {
441 my $self = shift;
442
443 if (@_ > 1) {
444 my $type = shift;
445
446 @_ = ($WH{$type} or Carp::croak "unsupported type passed to AnyEvent::Handle::push_write")
447 ->($self, @_);
448 }
449
450 if ($self->{filter_w}) {
451 $self->{filter_w}($self, \$_[0]);
452 } else {
453 $self->{wbuf} .= $_[0];
454 $self->_drain_wbuf;
455 }
456}
457
458=item $handle->push_write (type => @args)
459
460Instead of formatting your data yourself, you can also let this module do
461the job by specifying a type and type-specific arguments.
462
463Predefined types are (if you have ideas for additional types, feel free to
464drop by and tell us):
465
466=over 4
467
468=item netstring => $string
469
470Formats the given value as netstring
471(http://cr.yp.to/proto/netstrings.txt, this is not a recommendation to use them).
472
473=cut
474
475register_write_type netstring => sub {
476 my ($self, $string) = @_;
477
478 sprintf "%d:%s,", (length $string), $string
479};
480
481=item json => $array_or_hashref
482
483Encodes the given hash or array reference into a JSON object. Unless you
484provide your own JSON object, this means it will be encoded to JSON text
485in UTF-8.
486
487JSON objects (and arrays) are self-delimiting, so you can write JSON at
488one end of a handle and read them at the other end without using any
489additional framing.
490
491The generated JSON text is guaranteed not to contain any newlines: While
492this module doesn't need delimiters after or between JSON texts to be
493able to read them, many other languages depend on that.
494
495A simple RPC protocol that interoperates easily with others is to send
496JSON arrays (or objects, although arrays are usually the better choice as
497they mimic how function argument passing works) and a newline after each
498JSON text:
499
500 $handle->push_write (json => ["method", "arg1", "arg2"]); # whatever
501 $handle->push_write ("\012");
502
503An AnyEvent::Handle receiver would simply use the C<json> read type and
504rely on the fact that the newline will be skipped as leading whitespace:
505
506 $handle->push_read (json => sub { my $array = $_[1]; ... });
507
508Other languages could read single lines terminated by a newline and pass
509this line into their JSON decoder of choice.
510
511=cut
512
513register_write_type json => sub {
514 my ($self, $ref) = @_;
515
516 require JSON;
517
518 $self->{json} ? $self->{json}->encode ($ref)
519 : JSON::encode_json ($ref)
520};
521
522=back
523
524=item AnyEvent::Handle::register_write_type type => $coderef->($handle, @args)
525
526This function (not method) lets you add your own types to C<push_write>.
527Whenever the given C<type> is used, C<push_write> will invoke the code
528reference with the handle object and the remaining arguments.
529
530The code reference is supposed to return a single octet string that will
531be appended to the write buffer.
532
533Note that this is a function, and all types registered this way will be
534global, so try to use unique names.
535
536=cut
537
261############################################################################# 538#############################################################################
539
540=back
541
542=head2 READ QUEUE
543
544AnyEvent::Handle manages two queues per handle, one for writing and one
545for reading.
546
547The read queue is more complex than the write queue. It can be used in two
548ways, the "simple" way, using only C<on_read> and the "complex" way, using
549a queue.
550
551In the simple case, you just install an C<on_read> callback and whenever
552new data arrives, it will be called. You can then remove some data (if
553enough is there) from the read buffer (C<< $handle->rbuf >>) if you want
554or not.
555
556In the more complex case, you want to queue multiple callbacks. In this
557case, AnyEvent::Handle will call the first queued callback each time new
558data arrives and removes it when it has done its job (see C<push_read>,
559below).
560
561This way you can, for example, push three line-reads, followed by reading
562a chunk of data, and AnyEvent::Handle will execute them in order.
563
564Example 1: EPP protocol parser. EPP sends 4 byte length info, followed by
565the specified number of bytes which give an XML datagram.
566
567 # in the default state, expect some header bytes
568 $handle->on_read (sub {
569 # some data is here, now queue the length-header-read (4 octets)
570 shift->unshift_read (chunk => 4, sub {
571 # header arrived, decode
572 my $len = unpack "N", $_[1];
573
574 # now read the payload
575 shift->unshift_read (chunk => $len, sub {
576 my $xml = $_[1];
577 # handle xml
578 });
579 });
580 });
581
582Example 2: Implement a client for a protocol that replies either with
583"OK" and another line or "ERROR" for one request, and 64 bytes for the
584second request. Due tot he availability of a full queue, we can just
585pipeline sending both requests and manipulate the queue as necessary in
586the callbacks:
587
588 # request one
589 $handle->push_write ("request 1\015\012");
590
591 # we expect "ERROR" or "OK" as response, so push a line read
592 $handle->push_read (line => sub {
593 # if we got an "OK", we have to _prepend_ another line,
594 # so it will be read before the second request reads its 64 bytes
595 # which are already in the queue when this callback is called
596 # we don't do this in case we got an error
597 if ($_[1] eq "OK") {
598 $_[0]->unshift_read (line => sub {
599 my $response = $_[1];
600 ...
601 });
602 }
603 });
604
605 # request two
606 $handle->push_write ("request 2\015\012");
607
608 # simply read 64 bytes, always
609 $handle->push_read (chunk => 64, sub {
610 my $response = $_[1];
611 ...
612 });
613
614=over 4
615
616=cut
262 617
263sub _drain_rbuf { 618sub _drain_rbuf {
264 my ($self) = @_; 619 my ($self) = @_;
265 620
621 if (
622 defined $self->{rbuf_max}
623 && $self->{rbuf_max} < length $self->{rbuf}
624 ) {
625 return $self->_error (&Errno::ENOSPC, 1);
626 }
627
266 return if exists $self->{in_drain}; 628 return if $self->{in_drain};
267 local $self->{in_drain} = 1; 629 local $self->{in_drain} = 1;
268 630
269 while (my $len = length $self->{rbuf}) { 631 while (my $len = length $self->{rbuf}) {
270 no strict 'refs'; 632 no strict 'refs';
271 if (@{ $self->{queue} }) { 633 if (my $cb = shift @{ $self->{_queue} }) {
272 if ($self->{queue}[0]($self)) { 634 unless ($cb->($self)) {
273 shift @{ $self->{queue} };
274 } elsif ($self->{eof}) { 635 if ($self->{_eof}) {
275 # no progress can be made (not enough data and no data forthcoming) 636 # no progress can be made (not enough data and no data forthcoming)
276 $! = &Errno::EPIPE; return $self->error; 637 return $self->_error (&Errno::EPIPE, 1);
277 } else { 638 }
639
640 unshift @{ $self->{_queue} }, $cb;
278 return; 641 last;
279 } 642 }
280 } elsif ($self->{on_read}) { 643 } elsif ($self->{on_read}) {
281 $self->{on_read}($self); 644 $self->{on_read}($self);
282 645
283 if ( 646 if (
284 $self->{eof} # if no further data will arrive
285 && $len == length $self->{rbuf} # and no data has been consumed 647 $len == length $self->{rbuf} # if no data has been consumed
286 && !@{ $self->{queue} } # and the queue is still empty 648 && !@{ $self->{_queue} } # and the queue is still empty
287 && $self->{on_read} # and we still want to read data 649 && $self->{on_read} # but we still have on_read
288 ) { 650 ) {
651 # no further data will arrive
289 # then no progress can be made 652 # so no progress can be made
290 $! = &Errno::EPIPE; return $self->error; 653 return $self->_error (&Errno::EPIPE, 1)
654 if $self->{_eof};
655
656 last; # more data might arrive
291 } 657 }
292 } else { 658 } else {
293 # read side becomes idle 659 # read side becomes idle
294 delete $self->{rw}; 660 delete $self->{_rw};
295 return; 661 last;
296 } 662 }
297 } 663 }
298 664
299 if ($self->{eof}) {
300 $self->_shutdown;
301 $self->{on_eof}($self); 665 $self->{on_eof}($self)
666 if $self->{_eof} && $self->{on_eof};
667
668 # may need to restart read watcher
669 unless ($self->{_rw}) {
670 $self->start_read
671 if $self->{on_read} || @{ $self->{_queue} };
302 } 672 }
303} 673}
304 674
305=item $handle->on_read ($cb) 675=item $handle->on_read ($cb)
306 676
312 682
313sub on_read { 683sub on_read {
314 my ($self, $cb) = @_; 684 my ($self, $cb) = @_;
315 685
316 $self->{on_read} = $cb; 686 $self->{on_read} = $cb;
317
318 unless ($self->{rw} || $self->{eof}) {
319 Scalar::Util::weaken $self;
320
321 $self->{rw} = AnyEvent->io (fh => $self->{fh}, poll => "r", cb => sub {
322 my $len = sysread $self->{fh}, $self->{rbuf}, $self->{read_size} || 8192, length $self->{rbuf};
323
324 if ($len > 0) {
325 if (exists $self->{rbuf_max}) {
326 if ($self->{rbuf_max} < length $self->{rbuf}) {
327 $! = &Errno::ENOSPC; return $self->error;
328 }
329 }
330
331 } elsif (defined $len) {
332 $self->{eof} = 1;
333 delete $self->{rw};
334
335 } elsif ($! != EAGAIN && $! != EINTR) {
336 return $self->error;
337 }
338
339 $self->_drain_rbuf; 687 $self->_drain_rbuf if $cb;
340 });
341 }
342} 688}
343 689
344=item $handle->rbuf 690=item $handle->rbuf
345 691
346Returns the read buffer (as a modifiable lvalue). 692Returns the read buffer (as a modifiable lvalue).
365Append the given callback to the end of the queue (C<push_read>) or 711Append the given callback to the end of the queue (C<push_read>) or
366prepend it (C<unshift_read>). 712prepend it (C<unshift_read>).
367 713
368The callback is called each time some additional read data arrives. 714The callback is called each time some additional read data arrives.
369 715
370It must check wether enough data is in the read buffer already. 716It must check whether enough data is in the read buffer already.
371 717
372If not enough data is available, it must return the empty list or a false 718If not enough data is available, it must return the empty list or a false
373value, in which case it will be called repeatedly until enough data is 719value, in which case it will be called repeatedly until enough data is
374available (or an error condition is detected). 720available (or an error condition is detected).
375 721
377interested in (which can be none at all) and return a true value. After returning 723interested in (which can be none at all) and return a true value. After returning
378true, it will be removed from the queue. 724true, it will be removed from the queue.
379 725
380=cut 726=cut
381 727
728our %RH;
729
730sub register_read_type($$) {
731 $RH{$_[0]} = $_[1];
732}
733
382sub push_read { 734sub push_read {
383 my ($self, $cb) = @_; 735 my $self = shift;
736 my $cb = pop;
384 737
738 if (@_) {
739 my $type = shift;
740
741 $cb = ($RH{$type} or Carp::croak "unsupported type passed to AnyEvent::Handle::push_read")
742 ->($self, $cb, @_);
743 }
744
385 push @{ $self->{queue} }, $cb; 745 push @{ $self->{_queue} }, $cb;
386 $self->_drain_rbuf; 746 $self->_drain_rbuf;
387} 747}
388 748
389sub unshift_read { 749sub unshift_read {
390 my ($self, $cb) = @_; 750 my $self = shift;
751 my $cb = pop;
391 752
753 if (@_) {
754 my $type = shift;
755
756 $cb = ($RH{$type} or Carp::croak "unsupported type passed to AnyEvent::Handle::unshift_read")
757 ->($self, $cb, @_);
758 }
759
760
392 push @{ $self->{queue} }, $cb; 761 unshift @{ $self->{_queue} }, $cb;
393 $self->_drain_rbuf; 762 $self->_drain_rbuf;
394} 763}
395 764
396=item $handle->push_read_chunk ($len, $cb->($self, $data)) 765=item $handle->push_read (type => @args, $cb)
397 766
398=item $handle->unshift_read_chunk ($len, $cb->($self, $data)) 767=item $handle->unshift_read (type => @args, $cb)
399 768
400Append the given callback to the end of the queue (C<push_read_chunk>) or 769Instead of providing a callback that parses the data itself you can chose
401prepend it (C<unshift_read_chunk>). 770between a number of predefined parsing formats, for chunks of data, lines
771etc.
402 772
403The callback will be called only once C<$len> bytes have been read, and 773Predefined types are (if you have ideas for additional types, feel free to
404these C<$len> bytes will be passed to the callback. 774drop by and tell us):
405 775
406=cut 776=over 4
407 777
408sub _read_chunk($$) { 778=item chunk => $octets, $cb->($handle, $data)
409 my ($len, $cb) = @_; 779
780Invoke the callback only once C<$octets> bytes have been read. Pass the
781data read to the callback. The callback will never be called with less
782data.
783
784Example: read 2 bytes.
785
786 $handle->push_read (chunk => 2, sub {
787 warn "yay ", unpack "H*", $_[1];
788 });
789
790=cut
791
792register_read_type chunk => sub {
793 my ($self, $cb, $len) = @_;
410 794
411 sub { 795 sub {
412 $len <= length $_[0]{rbuf} or return; 796 $len <= length $_[0]{rbuf} or return;
413 $cb->($_[0], substr $_[0]{rbuf}, 0, $len, ""); 797 $cb->($_[0], substr $_[0]{rbuf}, 0, $len, "");
414 1 798 1
415 } 799 }
416} 800};
417 801
802# compatibility with older API
418sub push_read_chunk { 803sub push_read_chunk {
419 my ($self, $len, $cb) = @_; 804 $_[0]->push_read (chunk => $_[1], $_[2]);
420
421 $self->push_read (_read_chunk $len, $cb);
422} 805}
423
424 806
425sub unshift_read_chunk { 807sub unshift_read_chunk {
426 my ($self, $len, $cb) = @_; 808 $_[0]->unshift_read (chunk => $_[1], $_[2]);
427
428 $self->unshift_read (_read_chunk $len, $cb);
429} 809}
430 810
431=item $handle->push_read_line ([$eol, ]$cb->($self, $line, $eol)) 811=item line => [$eol, ]$cb->($handle, $line, $eol)
432
433=item $handle->unshift_read_line ([$eol, ]$cb->($self, $line, $eol))
434
435Append the given callback to the end of the queue (C<push_read_line>) or
436prepend it (C<unshift_read_line>).
437 812
438The callback will be called only once a full line (including the end of 813The callback will be called only once a full line (including the end of
439line marker, C<$eol>) has been read. This line (excluding the end of line 814line marker, C<$eol>) has been read. This line (excluding the end of line
440marker) will be passed to the callback as second argument (C<$line>), and 815marker) will be passed to the callback as second argument (C<$line>), and
441the end of line marker as the third argument (C<$eol>). 816the end of line marker as the third argument (C<$eol>).
452Partial lines at the end of the stream will never be returned, as they are 827Partial lines at the end of the stream will never be returned, as they are
453not marked by the end of line marker. 828not marked by the end of line marker.
454 829
455=cut 830=cut
456 831
457sub _read_line($$) { 832register_read_type line => sub {
458 my $cb = pop; 833 my ($self, $cb, $eol) = @_;
459 my $eol = @_ ? shift : qr|(\015?\012)|;
460 my $pos;
461 834
835 $eol = qr|(\015?\012)| if @_ < 3;
462 $eol = qr|(\Q$eol\E)| unless ref $eol; 836 $eol = quotemeta $eol unless ref $eol;
463 $eol = qr|^(.*?)($eol)|; 837 $eol = qr|^(.*?)($eol)|s;
464 838
465 sub { 839 sub {
466 $_[0]{rbuf} =~ s/$eol// or return; 840 $_[0]{rbuf} =~ s/$eol// or return;
467 841
468 $cb->($1, $2); 842 $cb->($_[0], $1, $2);
469 1 843 1
470 } 844 }
471} 845};
472 846
847# compatibility with older API
473sub push_read_line { 848sub push_read_line {
474 my $self = shift; 849 my $self = shift;
475
476 $self->push_read (&_read_line); 850 $self->push_read (line => @_);
477} 851}
478 852
479sub unshift_read_line { 853sub unshift_read_line {
480 my $self = shift; 854 my $self = shift;
481
482 $self->unshift_read (&_read_line); 855 $self->unshift_read (line => @_);
483} 856}
857
858=item netstring => $cb->($handle, $string)
859
860A netstring (http://cr.yp.to/proto/netstrings.txt, this is not an endorsement).
861
862Throws an error with C<$!> set to EBADMSG on format violations.
863
864=cut
865
866register_read_type netstring => sub {
867 my ($self, $cb) = @_;
868
869 sub {
870 unless ($_[0]{rbuf} =~ s/^(0|[1-9][0-9]*)://) {
871 if ($_[0]{rbuf} =~ /[^0-9]/) {
872 $self->_error (&Errno::EBADMSG);
873 }
874 return;
875 }
876
877 my $len = $1;
878
879 $self->unshift_read (chunk => $len, sub {
880 my $string = $_[1];
881 $_[0]->unshift_read (chunk => 1, sub {
882 if ($_[1] eq ",") {
883 $cb->($_[0], $string);
884 } else {
885 $self->_error (&Errno::EBADMSG);
886 }
887 });
888 });
889
890 1
891 }
892};
893
894=item regex => $accept[, $reject[, $skip], $cb->($handle, $data)
895
896Makes a regex match against the regex object C<$accept> and returns
897everything up to and including the match.
898
899Example: read a single line terminated by '\n'.
900
901 $handle->push_read (regex => qr<\n>, sub { ... });
902
903If C<$reject> is given and not undef, then it determines when the data is
904to be rejected: it is matched against the data when the C<$accept> regex
905does not match and generates an C<EBADMSG> error when it matches. This is
906useful to quickly reject wrong data (to avoid waiting for a timeout or a
907receive buffer overflow).
908
909Example: expect a single decimal number followed by whitespace, reject
910anything else (not the use of an anchor).
911
912 $handle->push_read (regex => qr<^[0-9]+\s>, qr<[^0-9]>, sub { ... });
913
914If C<$skip> is given and not C<undef>, then it will be matched against
915the receive buffer when neither C<$accept> nor C<$reject> match,
916and everything preceding and including the match will be accepted
917unconditionally. This is useful to skip large amounts of data that you
918know cannot be matched, so that the C<$accept> or C<$reject> regex do not
919have to start matching from the beginning. This is purely an optimisation
920and is usually worth only when you expect more than a few kilobytes.
921
922Example: expect a http header, which ends at C<\015\012\015\012>. Since we
923expect the header to be very large (it isn't in practise, but...), we use
924a skip regex to skip initial portions. The skip regex is tricky in that
925it only accepts something not ending in either \015 or \012, as these are
926required for the accept regex.
927
928 $handle->push_read (regex =>
929 qr<\015\012\015\012>,
930 undef, # no reject
931 qr<^.*[^\015\012]>,
932 sub { ... });
933
934=cut
935
936register_read_type regex => sub {
937 my ($self, $cb, $accept, $reject, $skip) = @_;
938
939 my $data;
940 my $rbuf = \$self->{rbuf};
941
942 sub {
943 # accept
944 if ($$rbuf =~ $accept) {
945 $data .= substr $$rbuf, 0, $+[0], "";
946 $cb->($self, $data);
947 return 1;
948 }
949
950 # reject
951 if ($reject && $$rbuf =~ $reject) {
952 $self->_error (&Errno::EBADMSG);
953 }
954
955 # skip
956 if ($skip && $$rbuf =~ $skip) {
957 $data .= substr $$rbuf, 0, $+[0], "";
958 }
959
960 ()
961 }
962};
963
964=item json => $cb->($handle, $hash_or_arrayref)
965
966Reads a JSON object or array, decodes it and passes it to the callback.
967
968If a C<json> object was passed to the constructor, then that will be used
969for the final decode, otherwise it will create a JSON coder expecting UTF-8.
970
971This read type uses the incremental parser available with JSON version
9722.09 (and JSON::XS version 2.2) and above. You have to provide a
973dependency on your own: this module will load the JSON module, but
974AnyEvent does not depend on it itself.
975
976Since JSON texts are fully self-delimiting, the C<json> read and write
977types are an ideal simple RPC protocol: just exchange JSON datagrams. See
978the C<json> write type description, above, for an actual example.
979
980=cut
981
982register_read_type json => sub {
983 my ($self, $cb, $accept, $reject, $skip) = @_;
984
985 require JSON;
986
987 my $data;
988 my $rbuf = \$self->{rbuf};
989
990 my $json = $self->{json} ||= JSON->new->utf8;
991
992 sub {
993 my $ref = $json->incr_parse ($self->{rbuf});
994
995 if ($ref) {
996 $self->{rbuf} = $json->incr_text;
997 $json->incr_text = "";
998 $cb->($self, $ref);
999
1000 1
1001 } else {
1002 $self->{rbuf} = "";
1003 ()
1004 }
1005 }
1006};
484 1007
485=back 1008=back
486 1009
1010=item AnyEvent::Handle::register_read_type type => $coderef->($handle, $cb, @args)
1011
1012This function (not method) lets you add your own types to C<push_read>.
1013
1014Whenever the given C<type> is used, C<push_read> will invoke the code
1015reference with the handle object, the callback and the remaining
1016arguments.
1017
1018The code reference is supposed to return a callback (usually a closure)
1019that works as a plain read callback (see C<< ->push_read ($cb) >>).
1020
1021It should invoke the passed callback when it is done reading (remember to
1022pass C<$handle> as first argument as all other callbacks do that).
1023
1024Note that this is a function, and all types registered this way will be
1025global, so try to use unique names.
1026
1027For examples, see the source of this module (F<perldoc -m AnyEvent::Handle>,
1028search for C<register_read_type>)).
1029
1030=item $handle->stop_read
1031
1032=item $handle->start_read
1033
1034In rare cases you actually do not want to read anything from the
1035socket. In this case you can call C<stop_read>. Neither C<on_read> nor
1036any queued callbacks will be executed then. To start reading again, call
1037C<start_read>.
1038
1039Note that AnyEvent::Handle will automatically C<start_read> for you when
1040you change the C<on_read> callback or push/unshift a read callback, and it
1041will automatically C<stop_read> for you when neither C<on_read> is set nor
1042there are any read requests in the queue.
1043
1044=cut
1045
1046sub stop_read {
1047 my ($self) = @_;
1048
1049 delete $self->{_rw};
1050}
1051
1052sub start_read {
1053 my ($self) = @_;
1054
1055 unless ($self->{_rw} || $self->{_eof}) {
1056 Scalar::Util::weaken $self;
1057
1058 $self->{_rw} = AnyEvent->io (fh => $self->{fh}, poll => "r", cb => sub {
1059 my $rbuf = $self->{filter_r} ? \my $buf : \$self->{rbuf};
1060 my $len = sysread $self->{fh}, $$rbuf, $self->{read_size} || 8192, length $$rbuf;
1061
1062 if ($len > 0) {
1063 $self->{_activity} = AnyEvent->now;
1064
1065 $self->{filter_r}
1066 ? $self->{filter_r}($self, $rbuf)
1067 : $self->_drain_rbuf;
1068
1069 } elsif (defined $len) {
1070 delete $self->{_rw};
1071 $self->{_eof} = 1;
1072 $self->_drain_rbuf;
1073
1074 } elsif ($! != EAGAIN && $! != EINTR && $! != WSAEWOULDBLOCK) {
1075 return $self->_error ($!, 1);
1076 }
1077 });
1078 }
1079}
1080
1081sub _dotls {
1082 my ($self) = @_;
1083
1084 my $buf;
1085
1086 if (length $self->{_tls_wbuf}) {
1087 while ((my $len = Net::SSLeay::write ($self->{tls}, $self->{_tls_wbuf})) > 0) {
1088 substr $self->{_tls_wbuf}, 0, $len, "";
1089 }
1090 }
1091
1092 if (length ($buf = Net::SSLeay::BIO_read ($self->{_wbio}))) {
1093 $self->{wbuf} .= $buf;
1094 $self->_drain_wbuf;
1095 }
1096
1097 while (defined ($buf = Net::SSLeay::read ($self->{tls}))) {
1098 if (length $buf) {
1099 $self->{rbuf} .= $buf;
1100 $self->_drain_rbuf;
1101 } else {
1102 # let's treat SSL-eof as we treat normal EOF
1103 $self->{_eof} = 1;
1104 $self->_shutdown;
1105 return;
1106 }
1107 }
1108
1109 my $err = Net::SSLeay::get_error ($self->{tls}, -1);
1110
1111 if ($err!= Net::SSLeay::ERROR_WANT_READ ()) {
1112 if ($err == Net::SSLeay::ERROR_SYSCALL ()) {
1113 return $self->_error ($!, 1);
1114 } elsif ($err == Net::SSLeay::ERROR_SSL ()) {
1115 return $self->_error (&Errno::EIO, 1);
1116 }
1117
1118 # all others are fine for our purposes
1119 }
1120}
1121
1122=item $handle->starttls ($tls[, $tls_ctx])
1123
1124Instead of starting TLS negotiation immediately when the AnyEvent::Handle
1125object is created, you can also do that at a later time by calling
1126C<starttls>.
1127
1128The first argument is the same as the C<tls> constructor argument (either
1129C<"connect">, C<"accept"> or an existing Net::SSLeay object).
1130
1131The second argument is the optional C<Net::SSLeay::CTX> object that is
1132used when AnyEvent::Handle has to create its own TLS connection object.
1133
1134The TLS connection object will end up in C<< $handle->{tls} >> after this
1135call and can be used or changed to your liking. Note that the handshake
1136might have already started when this function returns.
1137
1138=cut
1139
1140sub starttls {
1141 my ($self, $ssl, $ctx) = @_;
1142
1143 $self->stoptls;
1144
1145 if ($ssl eq "accept") {
1146 $ssl = Net::SSLeay::new ($ctx || TLS_CTX ());
1147 Net::SSLeay::set_accept_state ($ssl);
1148 } elsif ($ssl eq "connect") {
1149 $ssl = Net::SSLeay::new ($ctx || TLS_CTX ());
1150 Net::SSLeay::set_connect_state ($ssl);
1151 }
1152
1153 $self->{tls} = $ssl;
1154
1155 # basically, this is deep magic (because SSL_read should have the same issues)
1156 # but the openssl maintainers basically said: "trust us, it just works".
1157 # (unfortunately, we have to hardcode constants because the abysmally misdesigned
1158 # and mismaintained ssleay-module doesn't even offer them).
1159 # http://www.mail-archive.com/openssl-dev@openssl.org/msg22420.html
1160 Net::SSLeay::CTX_set_mode ($self->{tls},
1161 (eval { local $SIG{__DIE__}; Net::SSLeay::MODE_ENABLE_PARTIAL_WRITE () } || 1)
1162 | (eval { local $SIG{__DIE__}; Net::SSLeay::MODE_ACCEPT_MOVING_WRITE_BUFFER () } || 2));
1163
1164 $self->{_rbio} = Net::SSLeay::BIO_new (Net::SSLeay::BIO_s_mem ());
1165 $self->{_wbio} = Net::SSLeay::BIO_new (Net::SSLeay::BIO_s_mem ());
1166
1167 Net::SSLeay::set_bio ($ssl, $self->{_rbio}, $self->{_wbio});
1168
1169 $self->{filter_w} = sub {
1170 $_[0]{_tls_wbuf} .= ${$_[1]};
1171 &_dotls;
1172 };
1173 $self->{filter_r} = sub {
1174 Net::SSLeay::BIO_write ($_[0]{_rbio}, ${$_[1]});
1175 &_dotls;
1176 };
1177}
1178
1179=item $handle->stoptls
1180
1181Destroys the SSL connection, if any. Partial read or write data will be
1182lost.
1183
1184=cut
1185
1186sub stoptls {
1187 my ($self) = @_;
1188
1189 Net::SSLeay::free (delete $self->{tls}) if $self->{tls};
1190
1191 delete $self->{_rbio};
1192 delete $self->{_wbio};
1193 delete $self->{_tls_wbuf};
1194 delete $self->{filter_r};
1195 delete $self->{filter_w};
1196}
1197
1198sub DESTROY {
1199 my $self = shift;
1200
1201 $self->stoptls;
1202}
1203
1204=item AnyEvent::Handle::TLS_CTX
1205
1206This function creates and returns the Net::SSLeay::CTX object used by
1207default for TLS mode.
1208
1209The context is created like this:
1210
1211 Net::SSLeay::load_error_strings;
1212 Net::SSLeay::SSLeay_add_ssl_algorithms;
1213 Net::SSLeay::randomize;
1214
1215 my $CTX = Net::SSLeay::CTX_new;
1216
1217 Net::SSLeay::CTX_set_options $CTX, Net::SSLeay::OP_ALL
1218
1219=cut
1220
1221our $TLS_CTX;
1222
1223sub TLS_CTX() {
1224 $TLS_CTX || do {
1225 require Net::SSLeay;
1226
1227 Net::SSLeay::load_error_strings ();
1228 Net::SSLeay::SSLeay_add_ssl_algorithms ();
1229 Net::SSLeay::randomize ();
1230
1231 $TLS_CTX = Net::SSLeay::CTX_new ();
1232
1233 Net::SSLeay::CTX_set_options ($TLS_CTX, Net::SSLeay::OP_ALL ());
1234
1235 $TLS_CTX
1236 }
1237}
1238
1239=back
1240
1241=head1 SUBCLASSING AnyEvent::Handle
1242
1243In many cases, you might want to subclass AnyEvent::Handle.
1244
1245To make this easier, a given version of AnyEvent::Handle uses these
1246conventions:
1247
1248=over 4
1249
1250=item * all constructor arguments become object members.
1251
1252At least initially, when you pass a C<tls>-argument to the constructor it
1253will end up in C<< $handle->{tls} >>. Those members might be changes or
1254mutated later on (for example C<tls> will hold the TLS connection object).
1255
1256=item * other object member names are prefixed with an C<_>.
1257
1258All object members not explicitly documented (internal use) are prefixed
1259with an underscore character, so the remaining non-C<_>-namespace is free
1260for use for subclasses.
1261
1262=item * all members not documented here and not prefixed with an underscore
1263are free to use in subclasses.
1264
1265Of course, new versions of AnyEvent::Handle may introduce more "public"
1266member variables, but thats just life, at least it is documented.
1267
1268=back
1269
487=head1 AUTHOR 1270=head1 AUTHOR
488 1271
489Robin Redeker C<< <elmex at ta-sa.org> >>, Marc Lehmann <schmorp@schmorp.de>. 1272Robin Redeker C<< <elmex at ta-sa.org> >>, Marc Lehmann <schmorp@schmorp.de>.
490 1273
491=cut 1274=cut

Diff Legend

Removed lines
+ Added lines
< Changed lines
> Changed lines