1 |
elmex |
1.1 |
package AnyEvent::Handle; |
2 |
|
|
|
3 |
elmex |
1.6 |
no warnings; |
4 |
elmex |
1.1 |
use strict; |
5 |
|
|
|
6 |
root |
1.8 |
use AnyEvent (); |
7 |
|
|
use AnyEvent::Util (); |
8 |
|
|
use Scalar::Util (); |
9 |
|
|
use Carp (); |
10 |
|
|
use Fcntl (); |
11 |
elmex |
1.1 |
use Errno qw/EAGAIN EINTR/; |
12 |
|
|
|
13 |
|
|
=head1 NAME |
14 |
|
|
|
15 |
root |
1.22 |
AnyEvent::Handle - non-blocking I/O on file handles via AnyEvent |
16 |
elmex |
1.1 |
|
17 |
root |
1.15 |
This module is experimental. |
18 |
|
|
|
19 |
elmex |
1.1 |
=cut |
20 |
|
|
|
21 |
root |
1.15 |
our $VERSION = '0.04'; |
22 |
elmex |
1.1 |
|
23 |
|
|
=head1 SYNOPSIS |
24 |
|
|
|
25 |
|
|
use AnyEvent; |
26 |
|
|
use AnyEvent::Handle; |
27 |
|
|
|
28 |
|
|
my $cv = AnyEvent->condvar; |
29 |
|
|
|
30 |
|
|
my $ae_fh = AnyEvent::Handle->new (fh => \*STDIN); |
31 |
|
|
|
32 |
root |
1.8 |
#TODO |
33 |
elmex |
1.1 |
|
34 |
elmex |
1.2 |
# or use the constructor to pass the callback: |
35 |
|
|
|
36 |
|
|
my $ae_fh2 = |
37 |
|
|
AnyEvent::Handle->new ( |
38 |
|
|
fh => \*STDIN, |
39 |
|
|
on_eof => sub { |
40 |
|
|
$cv->broadcast; |
41 |
|
|
}, |
42 |
root |
1.8 |
#TODO |
43 |
elmex |
1.2 |
); |
44 |
|
|
|
45 |
elmex |
1.1 |
$cv->wait; |
46 |
|
|
|
47 |
|
|
=head1 DESCRIPTION |
48 |
|
|
|
49 |
root |
1.8 |
This module is a helper module to make it easier to do event-based I/O on |
50 |
elmex |
1.13 |
filehandles. For utility functions for doing non-blocking connects and accepts |
51 |
|
|
on sockets see L<AnyEvent::Util>. |
52 |
root |
1.8 |
|
53 |
|
|
In the following, when the documentation refers to of "bytes" then this |
54 |
|
|
means characters. As sysread and syswrite are used for all I/O, their |
55 |
|
|
treatment of characters applies to this module as well. |
56 |
elmex |
1.1 |
|
57 |
root |
1.8 |
All callbacks will be invoked with the handle object as their first |
58 |
|
|
argument. |
59 |
elmex |
1.1 |
|
60 |
|
|
=head1 METHODS |
61 |
|
|
|
62 |
|
|
=over 4 |
63 |
|
|
|
64 |
|
|
=item B<new (%args)> |
65 |
|
|
|
66 |
root |
1.8 |
The constructor supports these arguments (all as key => value pairs). |
67 |
elmex |
1.1 |
|
68 |
|
|
=over 4 |
69 |
|
|
|
70 |
root |
1.8 |
=item fh => $filehandle [MANDATORY] |
71 |
elmex |
1.1 |
|
72 |
|
|
The filehandle this L<AnyEvent::Handle> object will operate on. |
73 |
|
|
|
74 |
root |
1.8 |
NOTE: The filehandle will be set to non-blocking (using |
75 |
|
|
AnyEvent::Util::fh_nonblocking). |
76 |
|
|
|
77 |
root |
1.16 |
=item on_eof => $cb->($self) |
78 |
root |
1.10 |
|
79 |
|
|
Set the callback to be called on EOF. |
80 |
root |
1.8 |
|
81 |
root |
1.16 |
While not mandatory, it is highly recommended to set an eof callback, |
82 |
|
|
otherwise you might end up with a closed socket while you are still |
83 |
|
|
waiting for data. |
84 |
|
|
|
85 |
root |
1.10 |
=item on_error => $cb->($self) |
86 |
|
|
|
87 |
|
|
This is the fatal error callback, that is called when, well, a fatal error |
88 |
elmex |
1.20 |
occurs, such as not being able to resolve the hostname, failure to connect |
89 |
root |
1.10 |
or a read error. |
90 |
root |
1.8 |
|
91 |
|
|
The object will not be in a usable state when this callback has been |
92 |
|
|
called. |
93 |
|
|
|
94 |
root |
1.10 |
On callback entrance, the value of C<$!> contains the operating system |
95 |
root |
1.29 |
error (or C<ENOSPC>, C<EPIPE> or C<EBADMSG>). |
96 |
root |
1.8 |
|
97 |
root |
1.10 |
While not mandatory, it is I<highly> recommended to set this callback, as |
98 |
|
|
you will not be notified of errors otherwise. The default simply calls |
99 |
|
|
die. |
100 |
root |
1.8 |
|
101 |
|
|
=item on_read => $cb->($self) |
102 |
|
|
|
103 |
|
|
This sets the default read callback, which is called when data arrives |
104 |
root |
1.10 |
and no read request is in the queue. |
105 |
root |
1.8 |
|
106 |
|
|
To access (and remove data from) the read buffer, use the C<< ->rbuf >> |
107 |
elmex |
1.20 |
method or access the C<$self->{rbuf}> member directly. |
108 |
root |
1.8 |
|
109 |
|
|
When an EOF condition is detected then AnyEvent::Handle will first try to |
110 |
|
|
feed all the remaining data to the queued callbacks and C<on_read> before |
111 |
|
|
calling the C<on_eof> callback. If no progress can be made, then a fatal |
112 |
|
|
error will be raised (with C<$!> set to C<EPIPE>). |
113 |
elmex |
1.1 |
|
114 |
root |
1.8 |
=item on_drain => $cb->() |
115 |
elmex |
1.1 |
|
116 |
root |
1.8 |
This sets the callback that is called when the write buffer becomes empty |
117 |
|
|
(or when the callback is set and the buffer is empty already). |
118 |
elmex |
1.1 |
|
119 |
root |
1.8 |
To append to the write buffer, use the C<< ->push_write >> method. |
120 |
elmex |
1.2 |
|
121 |
root |
1.8 |
=item rbuf_max => <bytes> |
122 |
elmex |
1.2 |
|
123 |
root |
1.8 |
If defined, then a fatal error will be raised (with C<$!> set to C<ENOSPC>) |
124 |
|
|
when the read buffer ever (strictly) exceeds this size. This is useful to |
125 |
|
|
avoid denial-of-service attacks. |
126 |
elmex |
1.2 |
|
127 |
root |
1.8 |
For example, a server accepting connections from untrusted sources should |
128 |
|
|
be configured to accept only so-and-so much data that it cannot act on |
129 |
|
|
(for example, when expecting a line, an attacker could send an unlimited |
130 |
|
|
amount of data without a callback ever being called as long as the line |
131 |
|
|
isn't finished). |
132 |
elmex |
1.2 |
|
133 |
root |
1.8 |
=item read_size => <bytes> |
134 |
elmex |
1.2 |
|
135 |
root |
1.8 |
The default read block size (the amount of bytes this module will try to read |
136 |
|
|
on each [loop iteration). Default: C<4096>. |
137 |
|
|
|
138 |
|
|
=item low_water_mark => <bytes> |
139 |
|
|
|
140 |
|
|
Sets the amount of bytes (default: C<0>) that make up an "empty" write |
141 |
|
|
buffer: If the write reaches this size or gets even samller it is |
142 |
|
|
considered empty. |
143 |
elmex |
1.2 |
|
144 |
root |
1.19 |
=item tls => "accept" | "connect" | Net::SSLeay::SSL object |
145 |
|
|
|
146 |
|
|
When this parameter is given, it enables TLS (SSL) mode, that means it |
147 |
|
|
will start making tls handshake and will transparently encrypt/decrypt |
148 |
|
|
data. |
149 |
|
|
|
150 |
root |
1.26 |
TLS mode requires Net::SSLeay to be installed (it will be loaded |
151 |
|
|
automatically when you try to create a TLS handle). |
152 |
|
|
|
153 |
root |
1.19 |
For the TLS server side, use C<accept>, and for the TLS client side of a |
154 |
|
|
connection, use C<connect> mode. |
155 |
|
|
|
156 |
|
|
You can also provide your own TLS connection object, but you have |
157 |
|
|
to make sure that you call either C<Net::SSLeay::set_connect_state> |
158 |
|
|
or C<Net::SSLeay::set_accept_state> on it before you pass it to |
159 |
|
|
AnyEvent::Handle. |
160 |
|
|
|
161 |
root |
1.26 |
See the C<starttls> method if you need to start TLs negotiation later. |
162 |
|
|
|
163 |
root |
1.19 |
=item tls_ctx => $ssl_ctx |
164 |
|
|
|
165 |
|
|
Use the given Net::SSLeay::CTX object to create the new TLS connection |
166 |
|
|
(unless a connection object was specified directly). If this parameter is |
167 |
|
|
missing, then AnyEvent::Handle will use C<AnyEvent::Handle::TLS_CTX>. |
168 |
|
|
|
169 |
elmex |
1.1 |
=back |
170 |
|
|
|
171 |
|
|
=cut |
172 |
|
|
|
173 |
|
|
sub new { |
174 |
root |
1.8 |
my $class = shift; |
175 |
|
|
|
176 |
|
|
my $self = bless { @_ }, $class; |
177 |
|
|
|
178 |
|
|
$self->{fh} or Carp::croak "mandatory argument fh is missing"; |
179 |
|
|
|
180 |
|
|
AnyEvent::Util::fh_nonblocking $self->{fh}, 1; |
181 |
elmex |
1.1 |
|
182 |
root |
1.19 |
if ($self->{tls}) { |
183 |
|
|
require Net::SSLeay; |
184 |
|
|
$self->starttls (delete $self->{tls}, delete $self->{tls_ctx}); |
185 |
|
|
} |
186 |
|
|
|
187 |
root |
1.16 |
$self->on_eof (delete $self->{on_eof} ) if $self->{on_eof}; |
188 |
root |
1.10 |
$self->on_error (delete $self->{on_error}) if $self->{on_error}; |
189 |
root |
1.8 |
$self->on_drain (delete $self->{on_drain}) if $self->{on_drain}; |
190 |
|
|
$self->on_read (delete $self->{on_read} ) if $self->{on_read}; |
191 |
elmex |
1.1 |
|
192 |
root |
1.10 |
$self->start_read; |
193 |
|
|
|
194 |
root |
1.8 |
$self |
195 |
|
|
} |
196 |
elmex |
1.2 |
|
197 |
root |
1.8 |
sub _shutdown { |
198 |
|
|
my ($self) = @_; |
199 |
elmex |
1.2 |
|
200 |
root |
1.8 |
delete $self->{rw}; |
201 |
|
|
delete $self->{ww}; |
202 |
|
|
delete $self->{fh}; |
203 |
|
|
} |
204 |
|
|
|
205 |
|
|
sub error { |
206 |
|
|
my ($self) = @_; |
207 |
|
|
|
208 |
|
|
{ |
209 |
|
|
local $!; |
210 |
|
|
$self->_shutdown; |
211 |
elmex |
1.1 |
} |
212 |
|
|
|
213 |
root |
1.10 |
if ($self->{on_error}) { |
214 |
|
|
$self->{on_error}($self); |
215 |
|
|
} else { |
216 |
root |
1.29 |
Carp::croak "AnyEvent::Handle uncaught fatal error: $!"; |
217 |
root |
1.10 |
} |
218 |
elmex |
1.1 |
} |
219 |
|
|
|
220 |
root |
1.8 |
=item $fh = $handle->fh |
221 |
elmex |
1.1 |
|
222 |
root |
1.22 |
This method returns the file handle of the L<AnyEvent::Handle> object. |
223 |
elmex |
1.1 |
|
224 |
|
|
=cut |
225 |
|
|
|
226 |
|
|
sub fh { $_[0]->{fh} } |
227 |
|
|
|
228 |
root |
1.8 |
=item $handle->on_error ($cb) |
229 |
elmex |
1.1 |
|
230 |
root |
1.8 |
Replace the current C<on_error> callback (see the C<on_error> constructor argument). |
231 |
elmex |
1.1 |
|
232 |
root |
1.8 |
=cut |
233 |
|
|
|
234 |
|
|
sub on_error { |
235 |
|
|
$_[0]{on_error} = $_[1]; |
236 |
|
|
} |
237 |
|
|
|
238 |
|
|
=item $handle->on_eof ($cb) |
239 |
|
|
|
240 |
|
|
Replace the current C<on_eof> callback (see the C<on_eof> constructor argument). |
241 |
elmex |
1.1 |
|
242 |
|
|
=cut |
243 |
|
|
|
244 |
root |
1.8 |
sub on_eof { |
245 |
|
|
$_[0]{on_eof} = $_[1]; |
246 |
|
|
} |
247 |
|
|
|
248 |
root |
1.9 |
############################################################################# |
249 |
|
|
|
250 |
|
|
=back |
251 |
|
|
|
252 |
|
|
=head2 WRITE QUEUE |
253 |
|
|
|
254 |
|
|
AnyEvent::Handle manages two queues per handle, one for writing and one |
255 |
|
|
for reading. |
256 |
|
|
|
257 |
|
|
The write queue is very simple: you can add data to its end, and |
258 |
|
|
AnyEvent::Handle will automatically try to get rid of it for you. |
259 |
|
|
|
260 |
elmex |
1.20 |
When data could be written and the write buffer is shorter then the low |
261 |
root |
1.9 |
water mark, the C<on_drain> callback will be invoked. |
262 |
|
|
|
263 |
|
|
=over 4 |
264 |
|
|
|
265 |
root |
1.8 |
=item $handle->on_drain ($cb) |
266 |
|
|
|
267 |
|
|
Sets the C<on_drain> callback or clears it (see the description of |
268 |
|
|
C<on_drain> in the constructor). |
269 |
|
|
|
270 |
|
|
=cut |
271 |
|
|
|
272 |
|
|
sub on_drain { |
273 |
elmex |
1.1 |
my ($self, $cb) = @_; |
274 |
|
|
|
275 |
root |
1.8 |
$self->{on_drain} = $cb; |
276 |
|
|
|
277 |
|
|
$cb->($self) |
278 |
|
|
if $cb && $self->{low_water_mark} >= length $self->{wbuf}; |
279 |
|
|
} |
280 |
|
|
|
281 |
|
|
=item $handle->push_write ($data) |
282 |
|
|
|
283 |
|
|
Queues the given scalar to be written. You can push as much data as you |
284 |
|
|
want (only limited by the available memory), as C<AnyEvent::Handle> |
285 |
|
|
buffers it independently of the kernel. |
286 |
|
|
|
287 |
|
|
=cut |
288 |
|
|
|
289 |
root |
1.17 |
sub _drain_wbuf { |
290 |
|
|
my ($self) = @_; |
291 |
root |
1.8 |
|
292 |
root |
1.29 |
if (!$self->{ww} && length $self->{wbuf}) { |
293 |
root |
1.8 |
Scalar::Util::weaken $self; |
294 |
|
|
my $cb = sub { |
295 |
|
|
my $len = syswrite $self->{fh}, $self->{wbuf}; |
296 |
|
|
|
297 |
root |
1.29 |
if ($len >= 0) { |
298 |
root |
1.8 |
substr $self->{wbuf}, 0, $len, ""; |
299 |
|
|
|
300 |
|
|
$self->{on_drain}($self) |
301 |
|
|
if $self->{low_water_mark} >= length $self->{wbuf} |
302 |
|
|
&& $self->{on_drain}; |
303 |
|
|
|
304 |
|
|
delete $self->{ww} unless length $self->{wbuf}; |
305 |
|
|
} elsif ($! != EAGAIN && $! != EINTR) { |
306 |
|
|
$self->error; |
307 |
elmex |
1.1 |
} |
308 |
root |
1.8 |
}; |
309 |
|
|
|
310 |
|
|
$self->{ww} = AnyEvent->io (fh => $self->{fh}, poll => "w", cb => $cb); |
311 |
|
|
|
312 |
|
|
$cb->($self); |
313 |
|
|
}; |
314 |
|
|
} |
315 |
|
|
|
316 |
root |
1.30 |
our %WH; |
317 |
|
|
|
318 |
|
|
sub register_write_type($$) { |
319 |
|
|
$WH{$_[0]} = $_[1]; |
320 |
|
|
} |
321 |
|
|
|
322 |
root |
1.17 |
sub push_write { |
323 |
|
|
my $self = shift; |
324 |
|
|
|
325 |
root |
1.29 |
if (@_ > 1) { |
326 |
|
|
my $type = shift; |
327 |
|
|
|
328 |
|
|
@_ = ($WH{$type} or Carp::croak "unsupported type passed to AnyEvent::Handle::push_write") |
329 |
|
|
->($self, @_); |
330 |
|
|
} |
331 |
|
|
|
332 |
root |
1.17 |
if ($self->{filter_w}) { |
333 |
root |
1.18 |
$self->{filter_w}->($self, \$_[0]); |
334 |
root |
1.17 |
} else { |
335 |
|
|
$self->{wbuf} .= $_[0]; |
336 |
|
|
$self->_drain_wbuf; |
337 |
|
|
} |
338 |
|
|
} |
339 |
|
|
|
340 |
root |
1.29 |
=item $handle->push_write (type => @args) |
341 |
|
|
|
342 |
|
|
=item $handle->unshift_write (type => @args) |
343 |
|
|
|
344 |
|
|
Instead of formatting your data yourself, you can also let this module do |
345 |
|
|
the job by specifying a type and type-specific arguments. |
346 |
|
|
|
347 |
root |
1.30 |
Predefined types are (if you have ideas for additional types, feel free to |
348 |
|
|
drop by and tell us): |
349 |
root |
1.29 |
|
350 |
|
|
=over 4 |
351 |
|
|
|
352 |
|
|
=item netstring => $string |
353 |
|
|
|
354 |
|
|
Formats the given value as netstring |
355 |
|
|
(http://cr.yp.to/proto/netstrings.txt, this is not a recommendation to use them). |
356 |
|
|
|
357 |
root |
1.30 |
=back |
358 |
|
|
|
359 |
root |
1.29 |
=cut |
360 |
|
|
|
361 |
|
|
register_write_type netstring => sub { |
362 |
|
|
my ($self, $string) = @_; |
363 |
|
|
|
364 |
|
|
sprintf "%d:%s,", (length $string), $string |
365 |
|
|
}; |
366 |
|
|
|
367 |
root |
1.30 |
=item AnyEvent::Handle::register_write_type type => $coderef->($self, @args) |
368 |
|
|
|
369 |
|
|
This function (not method) lets you add your own types to C<push_write>. |
370 |
|
|
Whenever the given C<type> is used, C<push_write> will invoke the code |
371 |
|
|
reference with the handle object and the remaining arguments. |
372 |
root |
1.29 |
|
373 |
root |
1.30 |
The code reference is supposed to return a single octet string that will |
374 |
|
|
be appended to the write buffer. |
375 |
root |
1.29 |
|
376 |
root |
1.30 |
Note that this is a function, and all types registered this way will be |
377 |
|
|
global, so try to use unique names. |
378 |
root |
1.29 |
|
379 |
root |
1.30 |
=cut |
380 |
root |
1.29 |
|
381 |
root |
1.8 |
############################################################################# |
382 |
|
|
|
383 |
root |
1.9 |
=back |
384 |
|
|
|
385 |
|
|
=head2 READ QUEUE |
386 |
|
|
|
387 |
|
|
AnyEvent::Handle manages two queues per handle, one for writing and one |
388 |
|
|
for reading. |
389 |
|
|
|
390 |
|
|
The read queue is more complex than the write queue. It can be used in two |
391 |
|
|
ways, the "simple" way, using only C<on_read> and the "complex" way, using |
392 |
|
|
a queue. |
393 |
|
|
|
394 |
|
|
In the simple case, you just install an C<on_read> callback and whenever |
395 |
|
|
new data arrives, it will be called. You can then remove some data (if |
396 |
|
|
enough is there) from the read buffer (C<< $handle->rbuf >>) if you want |
397 |
|
|
or not. |
398 |
|
|
|
399 |
|
|
In the more complex case, you want to queue multiple callbacks. In this |
400 |
|
|
case, AnyEvent::Handle will call the first queued callback each time new |
401 |
|
|
data arrives and removes it when it has done its job (see C<push_read>, |
402 |
|
|
below). |
403 |
|
|
|
404 |
|
|
This way you can, for example, push three line-reads, followed by reading |
405 |
|
|
a chunk of data, and AnyEvent::Handle will execute them in order. |
406 |
|
|
|
407 |
|
|
Example 1: EPP protocol parser. EPP sends 4 byte length info, followed by |
408 |
|
|
the specified number of bytes which give an XML datagram. |
409 |
|
|
|
410 |
|
|
# in the default state, expect some header bytes |
411 |
|
|
$handle->on_read (sub { |
412 |
|
|
# some data is here, now queue the length-header-read (4 octets) |
413 |
|
|
shift->unshift_read_chunk (4, sub { |
414 |
|
|
# header arrived, decode |
415 |
|
|
my $len = unpack "N", $_[1]; |
416 |
|
|
|
417 |
|
|
# now read the payload |
418 |
|
|
shift->unshift_read_chunk ($len, sub { |
419 |
|
|
my $xml = $_[1]; |
420 |
|
|
# handle xml |
421 |
|
|
}); |
422 |
|
|
}); |
423 |
|
|
}); |
424 |
|
|
|
425 |
|
|
Example 2: Implement a client for a protocol that replies either with |
426 |
|
|
"OK" and another line or "ERROR" for one request, and 64 bytes for the |
427 |
|
|
second request. Due tot he availability of a full queue, we can just |
428 |
|
|
pipeline sending both requests and manipulate the queue as necessary in |
429 |
|
|
the callbacks: |
430 |
|
|
|
431 |
|
|
# request one |
432 |
|
|
$handle->push_write ("request 1\015\012"); |
433 |
|
|
|
434 |
|
|
# we expect "ERROR" or "OK" as response, so push a line read |
435 |
|
|
$handle->push_read_line (sub { |
436 |
|
|
# if we got an "OK", we have to _prepend_ another line, |
437 |
|
|
# so it will be read before the second request reads its 64 bytes |
438 |
|
|
# which are already in the queue when this callback is called |
439 |
|
|
# we don't do this in case we got an error |
440 |
|
|
if ($_[1] eq "OK") { |
441 |
|
|
$_[0]->unshift_read_line (sub { |
442 |
|
|
my $response = $_[1]; |
443 |
|
|
... |
444 |
|
|
}); |
445 |
|
|
} |
446 |
|
|
}); |
447 |
|
|
|
448 |
|
|
# request two |
449 |
|
|
$handle->push_write ("request 2\015\012"); |
450 |
|
|
|
451 |
|
|
# simply read 64 bytes, always |
452 |
|
|
$handle->push_read_chunk (64, sub { |
453 |
|
|
my $response = $_[1]; |
454 |
|
|
... |
455 |
|
|
}); |
456 |
|
|
|
457 |
|
|
=over 4 |
458 |
|
|
|
459 |
root |
1.10 |
=cut |
460 |
|
|
|
461 |
root |
1.8 |
sub _drain_rbuf { |
462 |
|
|
my ($self) = @_; |
463 |
elmex |
1.1 |
|
464 |
root |
1.17 |
if ( |
465 |
|
|
defined $self->{rbuf_max} |
466 |
|
|
&& $self->{rbuf_max} < length $self->{rbuf} |
467 |
|
|
) { |
468 |
|
|
$! = &Errno::ENOSPC; return $self->error; |
469 |
|
|
} |
470 |
|
|
|
471 |
root |
1.11 |
return if $self->{in_drain}; |
472 |
root |
1.8 |
local $self->{in_drain} = 1; |
473 |
elmex |
1.1 |
|
474 |
root |
1.8 |
while (my $len = length $self->{rbuf}) { |
475 |
|
|
no strict 'refs'; |
476 |
root |
1.10 |
if (my $cb = shift @{ $self->{queue} }) { |
477 |
root |
1.29 |
unless ($cb->($self)) { |
478 |
root |
1.10 |
if ($self->{eof}) { |
479 |
|
|
# no progress can be made (not enough data and no data forthcoming) |
480 |
|
|
$! = &Errno::EPIPE; return $self->error; |
481 |
|
|
} |
482 |
|
|
|
483 |
|
|
unshift @{ $self->{queue} }, $cb; |
484 |
root |
1.8 |
return; |
485 |
|
|
} |
486 |
|
|
} elsif ($self->{on_read}) { |
487 |
|
|
$self->{on_read}($self); |
488 |
|
|
|
489 |
|
|
if ( |
490 |
|
|
$self->{eof} # if no further data will arrive |
491 |
|
|
&& $len == length $self->{rbuf} # and no data has been consumed |
492 |
|
|
&& !@{ $self->{queue} } # and the queue is still empty |
493 |
|
|
&& $self->{on_read} # and we still want to read data |
494 |
|
|
) { |
495 |
|
|
# then no progress can be made |
496 |
|
|
$! = &Errno::EPIPE; return $self->error; |
497 |
elmex |
1.1 |
} |
498 |
root |
1.8 |
} else { |
499 |
|
|
# read side becomes idle |
500 |
|
|
delete $self->{rw}; |
501 |
|
|
return; |
502 |
|
|
} |
503 |
|
|
} |
504 |
|
|
|
505 |
|
|
if ($self->{eof}) { |
506 |
|
|
$self->_shutdown; |
507 |
root |
1.16 |
$self->{on_eof}($self) |
508 |
|
|
if $self->{on_eof}; |
509 |
root |
1.8 |
} |
510 |
elmex |
1.1 |
} |
511 |
|
|
|
512 |
root |
1.8 |
=item $handle->on_read ($cb) |
513 |
elmex |
1.1 |
|
514 |
root |
1.8 |
This replaces the currently set C<on_read> callback, or clears it (when |
515 |
|
|
the new callback is C<undef>). See the description of C<on_read> in the |
516 |
|
|
constructor. |
517 |
elmex |
1.1 |
|
518 |
root |
1.8 |
=cut |
519 |
|
|
|
520 |
|
|
sub on_read { |
521 |
|
|
my ($self, $cb) = @_; |
522 |
elmex |
1.1 |
|
523 |
root |
1.8 |
$self->{on_read} = $cb; |
524 |
elmex |
1.1 |
} |
525 |
|
|
|
526 |
root |
1.8 |
=item $handle->rbuf |
527 |
|
|
|
528 |
|
|
Returns the read buffer (as a modifiable lvalue). |
529 |
elmex |
1.1 |
|
530 |
root |
1.8 |
You can access the read buffer directly as the C<< ->{rbuf} >> member, if |
531 |
|
|
you want. |
532 |
elmex |
1.1 |
|
533 |
root |
1.8 |
NOTE: The read buffer should only be used or modified if the C<on_read>, |
534 |
|
|
C<push_read> or C<unshift_read> methods are used. The other read methods |
535 |
|
|
automatically manage the read buffer. |
536 |
elmex |
1.1 |
|
537 |
|
|
=cut |
538 |
|
|
|
539 |
elmex |
1.2 |
sub rbuf : lvalue { |
540 |
root |
1.8 |
$_[0]{rbuf} |
541 |
elmex |
1.2 |
} |
542 |
elmex |
1.1 |
|
543 |
root |
1.8 |
=item $handle->push_read ($cb) |
544 |
|
|
|
545 |
|
|
=item $handle->unshift_read ($cb) |
546 |
|
|
|
547 |
|
|
Append the given callback to the end of the queue (C<push_read>) or |
548 |
|
|
prepend it (C<unshift_read>). |
549 |
|
|
|
550 |
|
|
The callback is called each time some additional read data arrives. |
551 |
elmex |
1.1 |
|
552 |
elmex |
1.20 |
It must check whether enough data is in the read buffer already. |
553 |
elmex |
1.1 |
|
554 |
root |
1.8 |
If not enough data is available, it must return the empty list or a false |
555 |
|
|
value, in which case it will be called repeatedly until enough data is |
556 |
|
|
available (or an error condition is detected). |
557 |
|
|
|
558 |
|
|
If enough data was available, then the callback must remove all data it is |
559 |
|
|
interested in (which can be none at all) and return a true value. After returning |
560 |
|
|
true, it will be removed from the queue. |
561 |
elmex |
1.1 |
|
562 |
|
|
=cut |
563 |
|
|
|
564 |
root |
1.30 |
our %RH; |
565 |
|
|
|
566 |
|
|
sub register_read_type($$) { |
567 |
|
|
$RH{$_[0]} = $_[1]; |
568 |
|
|
} |
569 |
|
|
|
570 |
root |
1.8 |
sub push_read { |
571 |
root |
1.28 |
my $self = shift; |
572 |
|
|
my $cb = pop; |
573 |
|
|
|
574 |
|
|
if (@_) { |
575 |
|
|
my $type = shift; |
576 |
|
|
|
577 |
|
|
$cb = ($RH{$type} or Carp::croak "unsupported type passed to AnyEvent::Handle::push_read") |
578 |
|
|
->($self, $cb, @_); |
579 |
|
|
} |
580 |
elmex |
1.1 |
|
581 |
root |
1.8 |
push @{ $self->{queue} }, $cb; |
582 |
|
|
$self->_drain_rbuf; |
583 |
elmex |
1.1 |
} |
584 |
|
|
|
585 |
root |
1.8 |
sub unshift_read { |
586 |
root |
1.28 |
my $self = shift; |
587 |
|
|
my $cb = pop; |
588 |
|
|
|
589 |
|
|
if (@_) { |
590 |
|
|
my $type = shift; |
591 |
|
|
|
592 |
|
|
$cb = ($RH{$type} or Carp::croak "unsupported type passed to AnyEvent::Handle::unshift_read") |
593 |
|
|
->($self, $cb, @_); |
594 |
|
|
} |
595 |
|
|
|
596 |
root |
1.8 |
|
597 |
root |
1.28 |
unshift @{ $self->{queue} }, $cb; |
598 |
root |
1.8 |
$self->_drain_rbuf; |
599 |
|
|
} |
600 |
elmex |
1.1 |
|
601 |
root |
1.28 |
=item $handle->push_read (type => @args, $cb) |
602 |
elmex |
1.1 |
|
603 |
root |
1.28 |
=item $handle->unshift_read (type => @args, $cb) |
604 |
elmex |
1.1 |
|
605 |
root |
1.28 |
Instead of providing a callback that parses the data itself you can chose |
606 |
|
|
between a number of predefined parsing formats, for chunks of data, lines |
607 |
|
|
etc. |
608 |
elmex |
1.1 |
|
609 |
root |
1.30 |
Predefined types are (if you have ideas for additional types, feel free to |
610 |
|
|
drop by and tell us): |
611 |
root |
1.28 |
|
612 |
|
|
=over 4 |
613 |
|
|
|
614 |
|
|
=item chunk => $octets, $cb->($self, $data) |
615 |
|
|
|
616 |
|
|
Invoke the callback only once C<$octets> bytes have been read. Pass the |
617 |
|
|
data read to the callback. The callback will never be called with less |
618 |
|
|
data. |
619 |
|
|
|
620 |
|
|
Example: read 2 bytes. |
621 |
|
|
|
622 |
|
|
$handle->push_read (chunk => 2, sub { |
623 |
|
|
warn "yay ", unpack "H*", $_[1]; |
624 |
|
|
}); |
625 |
elmex |
1.1 |
|
626 |
|
|
=cut |
627 |
|
|
|
628 |
root |
1.28 |
register_read_type chunk => sub { |
629 |
|
|
my ($self, $cb, $len) = @_; |
630 |
elmex |
1.1 |
|
631 |
root |
1.8 |
sub { |
632 |
|
|
$len <= length $_[0]{rbuf} or return; |
633 |
elmex |
1.12 |
$cb->($_[0], substr $_[0]{rbuf}, 0, $len, ""); |
634 |
root |
1.8 |
1 |
635 |
|
|
} |
636 |
root |
1.28 |
}; |
637 |
root |
1.8 |
|
638 |
root |
1.28 |
# compatibility with older API |
639 |
root |
1.8 |
sub push_read_chunk { |
640 |
root |
1.28 |
$_[0]->push_read (chunk => $_[1], $_[2]); |
641 |
root |
1.8 |
} |
642 |
elmex |
1.1 |
|
643 |
root |
1.8 |
sub unshift_read_chunk { |
644 |
root |
1.28 |
$_[0]->unshift_read (chunk => $_[1], $_[2]); |
645 |
elmex |
1.1 |
} |
646 |
|
|
|
647 |
root |
1.28 |
=item line => [$eol, ]$cb->($self, $line, $eol) |
648 |
elmex |
1.1 |
|
649 |
root |
1.8 |
The callback will be called only once a full line (including the end of |
650 |
|
|
line marker, C<$eol>) has been read. This line (excluding the end of line |
651 |
|
|
marker) will be passed to the callback as second argument (C<$line>), and |
652 |
|
|
the end of line marker as the third argument (C<$eol>). |
653 |
elmex |
1.1 |
|
654 |
root |
1.8 |
The end of line marker, C<$eol>, can be either a string, in which case it |
655 |
|
|
will be interpreted as a fixed record end marker, or it can be a regex |
656 |
|
|
object (e.g. created by C<qr>), in which case it is interpreted as a |
657 |
|
|
regular expression. |
658 |
elmex |
1.1 |
|
659 |
root |
1.8 |
The end of line marker argument C<$eol> is optional, if it is missing (NOT |
660 |
|
|
undef), then C<qr|\015?\012|> is used (which is good for most internet |
661 |
|
|
protocols). |
662 |
elmex |
1.1 |
|
663 |
root |
1.8 |
Partial lines at the end of the stream will never be returned, as they are |
664 |
|
|
not marked by the end of line marker. |
665 |
elmex |
1.1 |
|
666 |
root |
1.8 |
=cut |
667 |
elmex |
1.1 |
|
668 |
root |
1.28 |
register_read_type line => sub { |
669 |
|
|
my ($self, $cb, $eol) = @_; |
670 |
elmex |
1.1 |
|
671 |
root |
1.28 |
$eol = qr|(\015?\012)| if @_ < 3; |
672 |
root |
1.14 |
$eol = quotemeta $eol unless ref $eol; |
673 |
|
|
$eol = qr|^(.*?)($eol)|s; |
674 |
elmex |
1.1 |
|
675 |
root |
1.8 |
sub { |
676 |
|
|
$_[0]{rbuf} =~ s/$eol// or return; |
677 |
elmex |
1.1 |
|
678 |
elmex |
1.12 |
$cb->($_[0], $1, $2); |
679 |
root |
1.8 |
1 |
680 |
|
|
} |
681 |
root |
1.28 |
}; |
682 |
elmex |
1.1 |
|
683 |
root |
1.28 |
# compatibility with older API |
684 |
root |
1.8 |
sub push_read_line { |
685 |
root |
1.28 |
my $self = shift; |
686 |
|
|
$self->push_read (line => @_); |
687 |
root |
1.10 |
} |
688 |
|
|
|
689 |
|
|
sub unshift_read_line { |
690 |
root |
1.28 |
my $self = shift; |
691 |
|
|
$self->unshift_read (line => @_); |
692 |
root |
1.10 |
} |
693 |
|
|
|
694 |
root |
1.29 |
=item netstring => $cb->($string) |
695 |
|
|
|
696 |
|
|
A netstring (http://cr.yp.to/proto/netstrings.txt, this is not an endorsement). |
697 |
|
|
|
698 |
|
|
Throws an error with C<$!> set to EBADMSG on format violations. |
699 |
|
|
|
700 |
|
|
=cut |
701 |
|
|
|
702 |
|
|
register_read_type netstring => sub { |
703 |
|
|
my ($self, $cb) = @_; |
704 |
|
|
|
705 |
|
|
sub { |
706 |
|
|
unless ($_[0]{rbuf} =~ s/^(0|[1-9][0-9]*)://) { |
707 |
|
|
if ($_[0]{rbuf} =~ /[^0-9]/) { |
708 |
|
|
$! = &Errno::EBADMSG; |
709 |
|
|
$self->error; |
710 |
|
|
} |
711 |
|
|
return; |
712 |
|
|
} |
713 |
|
|
|
714 |
|
|
my $len = $1; |
715 |
|
|
|
716 |
|
|
$self->unshift_read (chunk => $len, sub { |
717 |
|
|
my $string = $_[1]; |
718 |
|
|
$_[0]->unshift_read (chunk => 1, sub { |
719 |
|
|
if ($_[1] eq ",") { |
720 |
|
|
$cb->($_[0], $string); |
721 |
|
|
} else { |
722 |
|
|
$! = &Errno::EBADMSG; |
723 |
|
|
$self->error; |
724 |
|
|
} |
725 |
|
|
}); |
726 |
|
|
}); |
727 |
|
|
|
728 |
|
|
1 |
729 |
|
|
} |
730 |
|
|
}; |
731 |
|
|
|
732 |
root |
1.28 |
=back |
733 |
|
|
|
734 |
root |
1.30 |
=item AnyEvent::Handle::register_read_type type => $coderef->($self, $cb, @args) |
735 |
|
|
|
736 |
|
|
This function (not method) lets you add your own types to C<push_read>. |
737 |
|
|
|
738 |
|
|
Whenever the given C<type> is used, C<push_read> will invoke the code |
739 |
|
|
reference with the handle object, the callback and the remaining |
740 |
|
|
arguments. |
741 |
|
|
|
742 |
|
|
The code reference is supposed to return a callback (usually a closure) |
743 |
|
|
that works as a plain read callback (see C<< ->push_read ($cb) >>). |
744 |
|
|
|
745 |
|
|
It should invoke the passed callback when it is done reading (remember to |
746 |
|
|
pass C<$self> as first argument as all other callbacks do that). |
747 |
|
|
|
748 |
|
|
Note that this is a function, and all types registered this way will be |
749 |
|
|
global, so try to use unique names. |
750 |
|
|
|
751 |
|
|
For examples, see the source of this module (F<perldoc -m AnyEvent::Handle>, |
752 |
|
|
search for C<register_read_type>)). |
753 |
|
|
|
754 |
root |
1.10 |
=item $handle->stop_read |
755 |
|
|
|
756 |
|
|
=item $handle->start_read |
757 |
|
|
|
758 |
root |
1.18 |
In rare cases you actually do not want to read anything from the |
759 |
root |
1.10 |
socket. In this case you can call C<stop_read>. Neither C<on_read> no |
760 |
root |
1.22 |
any queued callbacks will be executed then. To start reading again, call |
761 |
root |
1.10 |
C<start_read>. |
762 |
|
|
|
763 |
|
|
=cut |
764 |
|
|
|
765 |
|
|
sub stop_read { |
766 |
|
|
my ($self) = @_; |
767 |
elmex |
1.1 |
|
768 |
root |
1.10 |
delete $self->{rw}; |
769 |
root |
1.8 |
} |
770 |
elmex |
1.1 |
|
771 |
root |
1.10 |
sub start_read { |
772 |
|
|
my ($self) = @_; |
773 |
|
|
|
774 |
|
|
unless ($self->{rw} || $self->{eof}) { |
775 |
|
|
Scalar::Util::weaken $self; |
776 |
|
|
|
777 |
|
|
$self->{rw} = AnyEvent->io (fh => $self->{fh}, poll => "r", cb => sub { |
778 |
root |
1.17 |
my $rbuf = $self->{filter_r} ? \my $buf : \$self->{rbuf}; |
779 |
|
|
my $len = sysread $self->{fh}, $$rbuf, $self->{read_size} || 8192, length $$rbuf; |
780 |
root |
1.10 |
|
781 |
|
|
if ($len > 0) { |
782 |
root |
1.17 |
$self->{filter_r} |
783 |
root |
1.18 |
? $self->{filter_r}->($self, $rbuf) |
784 |
root |
1.17 |
: $self->_drain_rbuf; |
785 |
root |
1.10 |
|
786 |
|
|
} elsif (defined $len) { |
787 |
root |
1.17 |
delete $self->{rw}; |
788 |
root |
1.10 |
$self->{eof} = 1; |
789 |
root |
1.17 |
$self->_drain_rbuf; |
790 |
root |
1.10 |
|
791 |
|
|
} elsif ($! != EAGAIN && $! != EINTR) { |
792 |
|
|
return $self->error; |
793 |
|
|
} |
794 |
|
|
}); |
795 |
|
|
} |
796 |
elmex |
1.1 |
} |
797 |
|
|
|
798 |
root |
1.19 |
sub _dotls { |
799 |
|
|
my ($self) = @_; |
800 |
|
|
|
801 |
|
|
if (length $self->{tls_wbuf}) { |
802 |
root |
1.22 |
while ((my $len = Net::SSLeay::write ($self->{tls}, $self->{tls_wbuf})) > 0) { |
803 |
|
|
substr $self->{tls_wbuf}, 0, $len, ""; |
804 |
|
|
} |
805 |
root |
1.19 |
} |
806 |
|
|
|
807 |
|
|
if (defined (my $buf = Net::SSLeay::BIO_read ($self->{tls_wbio}))) { |
808 |
|
|
$self->{wbuf} .= $buf; |
809 |
|
|
$self->_drain_wbuf; |
810 |
|
|
} |
811 |
|
|
|
812 |
root |
1.23 |
while (defined (my $buf = Net::SSLeay::read ($self->{tls}))) { |
813 |
|
|
$self->{rbuf} .= $buf; |
814 |
|
|
$self->_drain_rbuf; |
815 |
|
|
} |
816 |
|
|
|
817 |
root |
1.24 |
my $err = Net::SSLeay::get_error ($self->{tls}, -1); |
818 |
|
|
|
819 |
|
|
if ($err!= Net::SSLeay::ERROR_WANT_READ ()) { |
820 |
root |
1.23 |
if ($err == Net::SSLeay::ERROR_SYSCALL ()) { |
821 |
|
|
$self->error; |
822 |
|
|
} elsif ($err == Net::SSLeay::ERROR_SSL ()) { |
823 |
|
|
$! = &Errno::EIO; |
824 |
|
|
$self->error; |
825 |
root |
1.19 |
} |
826 |
root |
1.23 |
|
827 |
|
|
# all others are fine for our purposes |
828 |
root |
1.19 |
} |
829 |
|
|
} |
830 |
|
|
|
831 |
root |
1.25 |
=item $handle->starttls ($tls[, $tls_ctx]) |
832 |
|
|
|
833 |
|
|
Instead of starting TLS negotiation immediately when the AnyEvent::Handle |
834 |
|
|
object is created, you can also do that at a later time by calling |
835 |
|
|
C<starttls>. |
836 |
|
|
|
837 |
|
|
The first argument is the same as the C<tls> constructor argument (either |
838 |
|
|
C<"connect">, C<"accept"> or an existing Net::SSLeay object). |
839 |
|
|
|
840 |
|
|
The second argument is the optional C<Net::SSLeay::CTX> object that is |
841 |
|
|
used when AnyEvent::Handle has to create its own TLS connection object. |
842 |
|
|
|
843 |
|
|
=cut |
844 |
|
|
|
845 |
root |
1.19 |
# TODO: maybe document... |
846 |
|
|
sub starttls { |
847 |
|
|
my ($self, $ssl, $ctx) = @_; |
848 |
|
|
|
849 |
root |
1.25 |
$self->stoptls; |
850 |
|
|
|
851 |
root |
1.19 |
if ($ssl eq "accept") { |
852 |
|
|
$ssl = Net::SSLeay::new ($ctx || TLS_CTX ()); |
853 |
|
|
Net::SSLeay::set_accept_state ($ssl); |
854 |
|
|
} elsif ($ssl eq "connect") { |
855 |
|
|
$ssl = Net::SSLeay::new ($ctx || TLS_CTX ()); |
856 |
|
|
Net::SSLeay::set_connect_state ($ssl); |
857 |
|
|
} |
858 |
|
|
|
859 |
|
|
$self->{tls} = $ssl; |
860 |
|
|
|
861 |
root |
1.21 |
# basically, this is deep magic (because SSL_read should have the same issues) |
862 |
|
|
# but the openssl maintainers basically said: "trust us, it just works". |
863 |
|
|
# (unfortunately, we have to hardcode constants because the abysmally misdesigned |
864 |
|
|
# and mismaintained ssleay-module doesn't even offer them). |
865 |
root |
1.27 |
# http://www.mail-archive.com/openssl-dev@openssl.org/msg22420.html |
866 |
root |
1.21 |
Net::SSLeay::CTX_set_mode ($self->{tls}, |
867 |
|
|
(eval { Net::SSLeay::MODE_ENABLE_PARTIAL_WRITE () } || 1) |
868 |
|
|
| (eval { Net::SSLeay::MODE_ACCEPT_MOVING_WRITE_BUFFER () } || 2)); |
869 |
|
|
|
870 |
root |
1.19 |
$self->{tls_rbio} = Net::SSLeay::BIO_new (Net::SSLeay::BIO_s_mem ()); |
871 |
|
|
$self->{tls_wbio} = Net::SSLeay::BIO_new (Net::SSLeay::BIO_s_mem ()); |
872 |
|
|
|
873 |
|
|
Net::SSLeay::set_bio ($ssl, $self->{tls_rbio}, $self->{tls_wbio}); |
874 |
|
|
|
875 |
|
|
$self->{filter_w} = sub { |
876 |
|
|
$_[0]{tls_wbuf} .= ${$_[1]}; |
877 |
|
|
&_dotls; |
878 |
|
|
}; |
879 |
|
|
$self->{filter_r} = sub { |
880 |
|
|
Net::SSLeay::BIO_write ($_[0]{tls_rbio}, ${$_[1]}); |
881 |
|
|
&_dotls; |
882 |
|
|
}; |
883 |
|
|
} |
884 |
|
|
|
885 |
root |
1.25 |
=item $handle->stoptls |
886 |
|
|
|
887 |
|
|
Destroys the SSL connection, if any. Partial read or write data will be |
888 |
|
|
lost. |
889 |
|
|
|
890 |
|
|
=cut |
891 |
|
|
|
892 |
|
|
sub stoptls { |
893 |
|
|
my ($self) = @_; |
894 |
|
|
|
895 |
|
|
Net::SSLeay::free (delete $self->{tls}) if $self->{tls}; |
896 |
|
|
delete $self->{tls_rbio}; |
897 |
|
|
delete $self->{tls_wbio}; |
898 |
|
|
delete $self->{tls_wbuf}; |
899 |
|
|
delete $self->{filter_r}; |
900 |
|
|
delete $self->{filter_w}; |
901 |
|
|
} |
902 |
|
|
|
903 |
root |
1.19 |
sub DESTROY { |
904 |
|
|
my $self = shift; |
905 |
|
|
|
906 |
root |
1.25 |
$self->stoptls; |
907 |
root |
1.19 |
} |
908 |
|
|
|
909 |
|
|
=item AnyEvent::Handle::TLS_CTX |
910 |
|
|
|
911 |
|
|
This function creates and returns the Net::SSLeay::CTX object used by |
912 |
|
|
default for TLS mode. |
913 |
|
|
|
914 |
|
|
The context is created like this: |
915 |
|
|
|
916 |
|
|
Net::SSLeay::load_error_strings; |
917 |
|
|
Net::SSLeay::SSLeay_add_ssl_algorithms; |
918 |
|
|
Net::SSLeay::randomize; |
919 |
|
|
|
920 |
|
|
my $CTX = Net::SSLeay::CTX_new; |
921 |
|
|
|
922 |
|
|
Net::SSLeay::CTX_set_options $CTX, Net::SSLeay::OP_ALL |
923 |
|
|
|
924 |
|
|
=cut |
925 |
|
|
|
926 |
|
|
our $TLS_CTX; |
927 |
|
|
|
928 |
|
|
sub TLS_CTX() { |
929 |
|
|
$TLS_CTX || do { |
930 |
|
|
require Net::SSLeay; |
931 |
|
|
|
932 |
|
|
Net::SSLeay::load_error_strings (); |
933 |
|
|
Net::SSLeay::SSLeay_add_ssl_algorithms (); |
934 |
|
|
Net::SSLeay::randomize (); |
935 |
|
|
|
936 |
|
|
$TLS_CTX = Net::SSLeay::CTX_new (); |
937 |
|
|
|
938 |
|
|
Net::SSLeay::CTX_set_options ($TLS_CTX, Net::SSLeay::OP_ALL ()); |
939 |
|
|
|
940 |
|
|
$TLS_CTX |
941 |
|
|
} |
942 |
|
|
} |
943 |
|
|
|
944 |
elmex |
1.1 |
=back |
945 |
|
|
|
946 |
|
|
=head1 AUTHOR |
947 |
|
|
|
948 |
root |
1.8 |
Robin Redeker C<< <elmex at ta-sa.org> >>, Marc Lehmann <schmorp@schmorp.de>. |
949 |
elmex |
1.1 |
|
950 |
|
|
=cut |
951 |
|
|
|
952 |
|
|
1; # End of AnyEvent::Handle |