… | |
… | |
11 | |
11 | |
12 | my $hdl; $hdl = new AnyEvent::Handle |
12 | my $hdl; $hdl = new AnyEvent::Handle |
13 | fh => \*STDIN, |
13 | fh => \*STDIN, |
14 | on_error => sub { |
14 | on_error => sub { |
15 | my ($hdl, $fatal, $msg) = @_; |
15 | my ($hdl, $fatal, $msg) = @_; |
16 | warn "got error $msg\n"; |
16 | AE::log error => "got error $msg\n"; |
17 | $hdl->destroy; |
17 | $hdl->destroy; |
18 | $cv->send; |
18 | $cv->send; |
19 | }; |
19 | }; |
20 | |
20 | |
21 | # send some request line |
21 | # send some request line |
22 | $hdl->push_write ("getinfo\015\012"); |
22 | $hdl->push_write ("getinfo\015\012"); |
23 | |
23 | |
24 | # read the response line |
24 | # read the response line |
25 | $hdl->push_read (line => sub { |
25 | $hdl->push_read (line => sub { |
26 | my ($hdl, $line) = @_; |
26 | my ($hdl, $line) = @_; |
27 | warn "got line <$line>\n"; |
27 | say "got line <$line>"; |
28 | $cv->send; |
28 | $cv->send; |
29 | }); |
29 | }); |
30 | |
30 | |
31 | $cv->recv; |
31 | $cv->recv; |
32 | |
32 | |
… | |
… | |
75 | } |
75 | } |
76 | |
76 | |
77 | \&$func |
77 | \&$func |
78 | } |
78 | } |
79 | |
79 | |
|
|
80 | sub MAX_READ_SIZE() { 131072 } |
|
|
81 | |
80 | =head1 METHODS |
82 | =head1 METHODS |
81 | |
83 | |
82 | =over 4 |
84 | =over 4 |
83 | |
85 | |
84 | =item $handle = B<new> AnyEvent::Handle fh => $filehandle, key => value... |
86 | =item $handle = B<new> AnyEvent::Handle fh => $filehandle, key => value... |
… | |
… | |
112 | =over 4 |
114 | =over 4 |
113 | |
115 | |
114 | =item on_prepare => $cb->($handle) |
116 | =item on_prepare => $cb->($handle) |
115 | |
117 | |
116 | This (rarely used) callback is called before a new connection is |
118 | This (rarely used) callback is called before a new connection is |
117 | attempted, but after the file handle has been created. It could be used to |
119 | attempted, but after the file handle has been created (you can access that |
|
|
120 | file handle via C<< $handle->{fh} >>). It could be used to prepare the |
118 | prepare the file handle with parameters required for the actual connect |
121 | file handle with parameters required for the actual connect (as opposed to |
119 | (as opposed to settings that can be changed when the connection is already |
122 | settings that can be changed when the connection is already established). |
120 | established). |
|
|
121 | |
123 | |
122 | The return value of this callback should be the connect timeout value in |
124 | The return value of this callback should be the connect timeout value in |
123 | seconds (or C<0>, or C<undef>, or the empty list, to indicate that the |
125 | seconds (or C<0>, or C<undef>, or the empty list, to indicate that the |
124 | default timeout is to be used). |
126 | default timeout is to be used). |
125 | |
127 | |
126 | =item on_connect => $cb->($handle, $host, $port, $retry->()) |
128 | =item on_connect => $cb->($handle, $host, $port, $retry->()) |
127 | |
129 | |
128 | This callback is called when a connection has been successfully established. |
130 | This callback is called when a connection has been successfully established. |
129 | |
131 | |
130 | The peer's numeric host and port (the socket peername) are passed as |
132 | The peer's numeric host and port (the socket peername) are passed as |
131 | parameters, together with a retry callback. |
133 | parameters, together with a retry callback. At the time it is called the |
|
|
134 | read and write queues, EOF status, TLS status and similar properties of |
|
|
135 | the handle will have been reset. |
132 | |
136 | |
|
|
137 | It is not allowed to use the read or write queues while the handle object |
|
|
138 | is connecting. |
|
|
139 | |
133 | If, for some reason, the handle is not acceptable, calling C<$retry> |
140 | If, for some reason, the handle is not acceptable, calling C<$retry> will |
134 | will continue with the next connection target (in case of multi-homed |
141 | continue with the next connection target (in case of multi-homed hosts or |
135 | hosts or SRV records there can be multiple connection endpoints). At the |
142 | SRV records there can be multiple connection endpoints). The C<$retry> |
136 | time it is called the read and write queues, eof status, tls status and |
143 | callback can be invoked after the connect callback returns, i.e. one can |
137 | similar properties of the handle will have been reset. |
144 | start a handshake and then decide to retry with the next host if the |
|
|
145 | handshake fails. |
138 | |
146 | |
139 | In most cases, you should ignore the C<$retry> parameter. |
147 | In most cases, you should ignore the C<$retry> parameter. |
140 | |
148 | |
141 | =item on_connect_error => $cb->($handle, $message) |
149 | =item on_connect_error => $cb->($handle, $message) |
142 | |
150 | |
… | |
… | |
157 | |
165 | |
158 | Some errors are fatal (which is indicated by C<$fatal> being true). On |
166 | Some errors are fatal (which is indicated by C<$fatal> being true). On |
159 | fatal errors the handle object will be destroyed (by a call to C<< -> |
167 | fatal errors the handle object will be destroyed (by a call to C<< -> |
160 | destroy >>) after invoking the error callback (which means you are free to |
168 | destroy >>) after invoking the error callback (which means you are free to |
161 | examine the handle object). Examples of fatal errors are an EOF condition |
169 | examine the handle object). Examples of fatal errors are an EOF condition |
162 | with active (but unsatisifable) read watchers (C<EPIPE>) or I/O errors. In |
170 | with active (but unsatisfiable) read watchers (C<EPIPE>) or I/O errors. In |
163 | cases where the other side can close the connection at will, it is |
171 | cases where the other side can close the connection at will, it is |
164 | often easiest to not report C<EPIPE> errors in this callback. |
172 | often easiest to not report C<EPIPE> errors in this callback. |
165 | |
173 | |
166 | AnyEvent::Handle tries to find an appropriate error code for you to check |
174 | AnyEvent::Handle tries to find an appropriate error code for you to check |
167 | against, but in some cases (TLS errors), this does not work well. It is |
175 | against, but in some cases (TLS errors), this does not work well. It is |
… | |
… | |
222 | If an EOF condition has been detected but no C<on_eof> callback has been |
230 | If an EOF condition has been detected but no C<on_eof> callback has been |
223 | set, then a fatal error will be raised with C<$!> set to <0>. |
231 | set, then a fatal error will be raised with C<$!> set to <0>. |
224 | |
232 | |
225 | =item on_drain => $cb->($handle) |
233 | =item on_drain => $cb->($handle) |
226 | |
234 | |
227 | This sets the callback that is called when the write buffer becomes empty |
235 | This sets the callback that is called once when the write buffer becomes |
228 | (or immediately if the buffer is empty already). |
236 | empty (and immediately when the handle object is created). |
229 | |
237 | |
230 | To append to the write buffer, use the C<< ->push_write >> method. |
238 | To append to the write buffer, use the C<< ->push_write >> method. |
231 | |
239 | |
232 | This callback is useful when you don't want to put all of your write data |
240 | This callback is useful when you don't want to put all of your write data |
233 | into the queue at once, for example, when you want to write the contents |
241 | into the queue at once, for example, when you want to write the contents |
… | |
… | |
245 | many seconds pass without a successful read or write on the underlying |
253 | many seconds pass without a successful read or write on the underlying |
246 | file handle (or a call to C<timeout_reset>), the C<on_timeout> callback |
254 | file handle (or a call to C<timeout_reset>), the C<on_timeout> callback |
247 | will be invoked (and if that one is missing, a non-fatal C<ETIMEDOUT> |
255 | will be invoked (and if that one is missing, a non-fatal C<ETIMEDOUT> |
248 | error will be raised). |
256 | error will be raised). |
249 | |
257 | |
250 | There are three variants of the timeouts that work independently |
258 | There are three variants of the timeouts that work independently of each |
251 | of each other, for both read and write, just read, and just write: |
259 | other, for both read and write (triggered when nothing was read I<OR> |
|
|
260 | written), just read (triggered when nothing was read), and just write: |
252 | C<timeout>, C<rtimeout> and C<wtimeout>, with corresponding callbacks |
261 | C<timeout>, C<rtimeout> and C<wtimeout>, with corresponding callbacks |
253 | C<on_timeout>, C<on_rtimeout> and C<on_wtimeout>, and reset functions |
262 | C<on_timeout>, C<on_rtimeout> and C<on_wtimeout>, and reset functions |
254 | C<timeout_reset>, C<rtimeout_reset>, and C<wtimeout_reset>. |
263 | C<timeout_reset>, C<rtimeout_reset>, and C<wtimeout_reset>. |
255 | |
264 | |
256 | Note that timeout processing is active even when you do not have |
265 | Note that timeout processing is active even when you do not have any |
257 | any outstanding read or write requests: If you plan to keep the connection |
266 | outstanding read or write requests: If you plan to keep the connection |
258 | idle then you should disable the timeout temporarily or ignore the timeout |
267 | idle then you should disable the timeout temporarily or ignore the |
259 | in the C<on_timeout> callback, in which case AnyEvent::Handle will simply |
268 | timeout in the corresponding C<on_timeout> callback, in which case |
260 | restart the timeout. |
269 | AnyEvent::Handle will simply restart the timeout. |
261 | |
270 | |
262 | Zero (the default) disables this timeout. |
271 | Zero (the default) disables the corresponding timeout. |
263 | |
272 | |
264 | =item on_timeout => $cb->($handle) |
273 | =item on_timeout => $cb->($handle) |
|
|
274 | |
|
|
275 | =item on_rtimeout => $cb->($handle) |
|
|
276 | |
|
|
277 | =item on_wtimeout => $cb->($handle) |
265 | |
278 | |
266 | Called whenever the inactivity timeout passes. If you return from this |
279 | Called whenever the inactivity timeout passes. If you return from this |
267 | callback, then the timeout will be reset as if some activity had happened, |
280 | callback, then the timeout will be reset as if some activity had happened, |
268 | so this condition is not fatal in any way. |
281 | so this condition is not fatal in any way. |
269 | |
282 | |
… | |
… | |
276 | For example, a server accepting connections from untrusted sources should |
289 | For example, a server accepting connections from untrusted sources should |
277 | be configured to accept only so-and-so much data that it cannot act on |
290 | be configured to accept only so-and-so much data that it cannot act on |
278 | (for example, when expecting a line, an attacker could send an unlimited |
291 | (for example, when expecting a line, an attacker could send an unlimited |
279 | amount of data without a callback ever being called as long as the line |
292 | amount of data without a callback ever being called as long as the line |
280 | isn't finished). |
293 | isn't finished). |
|
|
294 | |
|
|
295 | =item wbuf_max => <bytes> |
|
|
296 | |
|
|
297 | If defined, then a fatal error will be raised (with C<$!> set to C<ENOSPC>) |
|
|
298 | when the write buffer ever (strictly) exceeds this size. This is useful to |
|
|
299 | avoid some forms of denial-of-service attacks. |
|
|
300 | |
|
|
301 | Although the units of this parameter is bytes, this is the I<raw> number |
|
|
302 | of bytes not yet accepted by the kernel. This can make a difference when |
|
|
303 | you e.g. use TLS, as TLS typically makes your write data larger (but it |
|
|
304 | can also make it smaller due to compression). |
|
|
305 | |
|
|
306 | As an example of when this limit is useful, take a chat server that sends |
|
|
307 | chat messages to a client. If the client does not read those in a timely |
|
|
308 | manner then the send buffer in the server would grow unbounded. |
281 | |
309 | |
282 | =item autocork => <boolean> |
310 | =item autocork => <boolean> |
283 | |
311 | |
284 | When disabled (the default), C<push_write> will try to immediately |
312 | When disabled (the default), C<push_write> will try to immediately |
285 | write the data to the handle if possible. This avoids having to register |
313 | write the data to the handle if possible. This avoids having to register |
… | |
… | |
337 | already have occured on BSD systems), but at least it will protect you |
365 | already have occured on BSD systems), but at least it will protect you |
338 | from most attacks. |
366 | from most attacks. |
339 | |
367 | |
340 | =item read_size => <bytes> |
368 | =item read_size => <bytes> |
341 | |
369 | |
342 | The default read block size (the number of bytes this module will |
370 | The initial read block size, the number of bytes this module will try |
343 | try to read during each loop iteration, which affects memory |
371 | to read during each loop iteration. Each handle object will consume |
344 | requirements). Default: C<8192>. |
372 | at least this amount of memory for the read buffer as well, so when |
|
|
373 | handling many connections watch out for memory requirements). See also |
|
|
374 | C<max_read_size>. Default: C<2048>. |
|
|
375 | |
|
|
376 | =item max_read_size => <bytes> |
|
|
377 | |
|
|
378 | The maximum read buffer size used by the dynamic adjustment |
|
|
379 | algorithm: Each time AnyEvent::Handle can read C<read_size> bytes in |
|
|
380 | one go it will double C<read_size> up to the maximum given by this |
|
|
381 | option. Default: C<131072> or C<read_size>, whichever is higher. |
345 | |
382 | |
346 | =item low_water_mark => <bytes> |
383 | =item low_water_mark => <bytes> |
347 | |
384 | |
348 | Sets the number of bytes (default: C<0>) that make up an "empty" write |
385 | Sets the number of bytes (default: C<0>) that make up an "empty" write |
349 | buffer: If the buffer reaches this size or gets even samller it is |
386 | buffer: If the buffer reaches this size or gets even samller it is |
… | |
… | |
412 | Use the C<< ->starttls >> method if you need to start TLS negotiation later. |
449 | Use the C<< ->starttls >> method if you need to start TLS negotiation later. |
413 | |
450 | |
414 | =item tls_ctx => $anyevent_tls |
451 | =item tls_ctx => $anyevent_tls |
415 | |
452 | |
416 | Use the given C<AnyEvent::TLS> object to create the new TLS connection |
453 | Use the given C<AnyEvent::TLS> object to create the new TLS connection |
417 | (unless a connection object was specified directly). If this parameter is |
454 | (unless a connection object was specified directly). If this |
418 | missing, then AnyEvent::Handle will use C<AnyEvent::Handle::TLS_CTX>. |
455 | parameter is missing (or C<undef>), then AnyEvent::Handle will use |
|
|
456 | C<AnyEvent::Handle::TLS_CTX>. |
419 | |
457 | |
420 | Instead of an object, you can also specify a hash reference with C<< key |
458 | Instead of an object, you can also specify a hash reference with C<< key |
421 | => value >> pairs. Those will be passed to L<AnyEvent::TLS> to create a |
459 | => value >> pairs. Those will be passed to L<AnyEvent::TLS> to create a |
422 | new TLS context object. |
460 | new TLS context object. |
423 | |
461 | |
… | |
… | |
492 | $self->{connect}[0], |
530 | $self->{connect}[0], |
493 | $self->{connect}[1], |
531 | $self->{connect}[1], |
494 | sub { |
532 | sub { |
495 | my ($fh, $host, $port, $retry) = @_; |
533 | my ($fh, $host, $port, $retry) = @_; |
496 | |
534 | |
|
|
535 | delete $self->{_connect}; # no longer needed |
|
|
536 | |
497 | if ($fh) { |
537 | if ($fh) { |
498 | $self->{fh} = $fh; |
538 | $self->{fh} = $fh; |
499 | |
539 | |
500 | delete $self->{_skip_drain_rbuf}; |
540 | delete $self->{_skip_drain_rbuf}; |
501 | $self->_start; |
541 | $self->_start; |
… | |
… | |
508 | }); |
548 | }); |
509 | |
549 | |
510 | } else { |
550 | } else { |
511 | if ($self->{on_connect_error}) { |
551 | if ($self->{on_connect_error}) { |
512 | $self->{on_connect_error}($self, "$!"); |
552 | $self->{on_connect_error}($self, "$!"); |
513 | $self->destroy; |
553 | $self->destroy if $self; |
514 | } else { |
554 | } else { |
515 | $self->_error ($!, 1); |
555 | $self->_error ($!, 1); |
516 | } |
556 | } |
517 | } |
557 | } |
518 | }, |
558 | }, |
519 | sub { |
559 | sub { |
520 | local $self->{fh} = $_[0]; |
560 | local $self->{fh} = $_[0]; |
521 | |
561 | |
522 | $self->{on_prepare} |
562 | $self->{on_prepare} |
523 | ? $self->{on_prepare}->($self) |
563 | ? $self->{on_prepare}->($self) |
524 | : () |
564 | : () |
525 | } |
565 | } |
526 | ); |
566 | ); |
527 | } |
567 | } |
528 | |
568 | |
… | |
… | |
545 | AnyEvent::Util::fh_nonblocking $self->{fh}, 1; |
585 | AnyEvent::Util::fh_nonblocking $self->{fh}, 1; |
546 | |
586 | |
547 | $self->{_activity} = |
587 | $self->{_activity} = |
548 | $self->{_ractivity} = |
588 | $self->{_ractivity} = |
549 | $self->{_wactivity} = AE::now; |
589 | $self->{_wactivity} = AE::now; |
|
|
590 | |
|
|
591 | $self->{read_size} ||= 2048; |
|
|
592 | $self->{max_read_size} = $self->{read_size} |
|
|
593 | if $self->{read_size} > ($self->{max_read_size} || MAX_READ_SIZE); |
550 | |
594 | |
551 | $self->timeout (delete $self->{timeout} ) if $self->{timeout}; |
595 | $self->timeout (delete $self->{timeout} ) if $self->{timeout}; |
552 | $self->rtimeout (delete $self->{rtimeout} ) if $self->{rtimeout}; |
596 | $self->rtimeout (delete $self->{rtimeout} ) if $self->{rtimeout}; |
553 | $self->wtimeout (delete $self->{wtimeout} ) if $self->{wtimeout}; |
597 | $self->wtimeout (delete $self->{wtimeout} ) if $self->{wtimeout}; |
554 | |
598 | |
… | |
… | |
723 | |
767 | |
724 | =item $handle->rbuf_max ($max_octets) |
768 | =item $handle->rbuf_max ($max_octets) |
725 | |
769 | |
726 | Configures the C<rbuf_max> setting (C<undef> disables it). |
770 | Configures the C<rbuf_max> setting (C<undef> disables it). |
727 | |
771 | |
|
|
772 | =item $handle->wbuf_max ($max_octets) |
|
|
773 | |
|
|
774 | Configures the C<wbuf_max> setting (C<undef> disables it). |
|
|
775 | |
728 | =cut |
776 | =cut |
729 | |
777 | |
730 | sub rbuf_max { |
778 | sub rbuf_max { |
731 | $_[0]{rbuf_max} = $_[1]; |
779 | $_[0]{rbuf_max} = $_[1]; |
732 | } |
780 | } |
733 | |
781 | |
|
|
782 | sub wbuf_max { |
|
|
783 | $_[0]{wbuf_max} = $_[1]; |
|
|
784 | } |
|
|
785 | |
734 | ############################################################################# |
786 | ############################################################################# |
735 | |
787 | |
736 | =item $handle->timeout ($seconds) |
788 | =item $handle->timeout ($seconds) |
737 | |
789 | |
738 | =item $handle->rtimeout ($seconds) |
790 | =item $handle->rtimeout ($seconds) |
739 | |
791 | |
740 | =item $handle->wtimeout ($seconds) |
792 | =item $handle->wtimeout ($seconds) |
741 | |
793 | |
742 | Configures (or disables) the inactivity timeout. |
794 | Configures (or disables) the inactivity timeout. |
|
|
795 | |
|
|
796 | The timeout will be checked instantly, so this method might destroy the |
|
|
797 | handle before it returns. |
743 | |
798 | |
744 | =item $handle->timeout_reset |
799 | =item $handle->timeout_reset |
745 | |
800 | |
746 | =item $handle->rtimeout_reset |
801 | =item $handle->rtimeout_reset |
747 | |
802 | |
… | |
… | |
831 | |
886 | |
832 | The write queue is very simple: you can add data to its end, and |
887 | The write queue is very simple: you can add data to its end, and |
833 | AnyEvent::Handle will automatically try to get rid of it for you. |
888 | AnyEvent::Handle will automatically try to get rid of it for you. |
834 | |
889 | |
835 | When data could be written and the write buffer is shorter then the low |
890 | When data could be written and the write buffer is shorter then the low |
836 | water mark, the C<on_drain> callback will be invoked. |
891 | water mark, the C<on_drain> callback will be invoked once. |
837 | |
892 | |
838 | =over 4 |
893 | =over 4 |
839 | |
894 | |
840 | =item $handle->on_drain ($cb) |
895 | =item $handle->on_drain ($cb) |
841 | |
896 | |
… | |
… | |
856 | if $cb && $self->{low_water_mark} >= (length $self->{wbuf}) + (length $self->{_tls_wbuf}); |
911 | if $cb && $self->{low_water_mark} >= (length $self->{wbuf}) + (length $self->{_tls_wbuf}); |
857 | } |
912 | } |
858 | |
913 | |
859 | =item $handle->push_write ($data) |
914 | =item $handle->push_write ($data) |
860 | |
915 | |
861 | Queues the given scalar to be written. You can push as much data as you |
916 | Queues the given scalar to be written. You can push as much data as |
862 | want (only limited by the available memory), as C<AnyEvent::Handle> |
917 | you want (only limited by the available memory and C<wbuf_max>), as |
863 | buffers it independently of the kernel. |
918 | C<AnyEvent::Handle> buffers it independently of the kernel. |
864 | |
919 | |
865 | This method may invoke callbacks (and therefore the handle might be |
920 | This method may invoke callbacks (and therefore the handle might be |
866 | destroyed after it returns). |
921 | destroyed after it returns). |
867 | |
922 | |
868 | =cut |
923 | =cut |
… | |
… | |
896 | $cb->() unless $self->{autocork}; |
951 | $cb->() unless $self->{autocork}; |
897 | |
952 | |
898 | # if still data left in wbuf, we need to poll |
953 | # if still data left in wbuf, we need to poll |
899 | $self->{_ww} = AE::io $self->{fh}, 1, $cb |
954 | $self->{_ww} = AE::io $self->{fh}, 1, $cb |
900 | if length $self->{wbuf}; |
955 | if length $self->{wbuf}; |
|
|
956 | |
|
|
957 | if ( |
|
|
958 | defined $self->{wbuf_max} |
|
|
959 | && $self->{wbuf_max} < length $self->{wbuf} |
|
|
960 | ) { |
|
|
961 | $self->_error (Errno::ENOSPC, 1), return; |
|
|
962 | } |
901 | }; |
963 | }; |
902 | } |
964 | } |
903 | |
965 | |
904 | our %WH; |
966 | our %WH; |
905 | |
967 | |
… | |
… | |
1025 | =cut |
1087 | =cut |
1026 | |
1088 | |
1027 | register_write_type storable => sub { |
1089 | register_write_type storable => sub { |
1028 | my ($self, $ref) = @_; |
1090 | my ($self, $ref) = @_; |
1029 | |
1091 | |
1030 | require Storable; |
1092 | require Storable unless $Storable::VERSION; |
1031 | |
1093 | |
1032 | pack "w/a*", Storable::nfreeze ($ref) |
1094 | pack "w/a*", Storable::nfreeze ($ref) |
1033 | }; |
1095 | }; |
1034 | |
1096 | |
1035 | =back |
1097 | =back |
… | |
… | |
1040 | before it was actually written. One way to do that is to replace your |
1102 | before it was actually written. One way to do that is to replace your |
1041 | C<on_drain> handler by a callback that shuts down the socket (and set |
1103 | C<on_drain> handler by a callback that shuts down the socket (and set |
1042 | C<low_water_mark> to C<0>). This method is a shorthand for just that, and |
1104 | C<low_water_mark> to C<0>). This method is a shorthand for just that, and |
1043 | replaces the C<on_drain> callback with: |
1105 | replaces the C<on_drain> callback with: |
1044 | |
1106 | |
1045 | sub { shutdown $_[0]{fh}, 1 } # for push_shutdown |
1107 | sub { shutdown $_[0]{fh}, 1 } |
1046 | |
1108 | |
1047 | This simply shuts down the write side and signals an EOF condition to the |
1109 | This simply shuts down the write side and signals an EOF condition to the |
1048 | the peer. |
1110 | the peer. |
1049 | |
1111 | |
1050 | You can rely on the normal read queue and C<on_eof> handling |
1112 | You can rely on the normal read queue and C<on_eof> handling |
… | |
… | |
1072 | |
1134 | |
1073 | Whenever the given C<type> is used, C<push_write> will the function with |
1135 | Whenever the given C<type> is used, C<push_write> will the function with |
1074 | the handle object and the remaining arguments. |
1136 | the handle object and the remaining arguments. |
1075 | |
1137 | |
1076 | The function is supposed to return a single octet string that will be |
1138 | The function is supposed to return a single octet string that will be |
1077 | appended to the write buffer, so you cna mentally treat this function as a |
1139 | appended to the write buffer, so you can mentally treat this function as a |
1078 | "arguments to on-the-wire-format" converter. |
1140 | "arguments to on-the-wire-format" converter. |
1079 | |
1141 | |
1080 | Example: implement a custom write type C<join> that joins the remaining |
1142 | Example: implement a custom write type C<join> that joins the remaining |
1081 | arguments using the first one. |
1143 | arguments using the first one. |
1082 | |
1144 | |
… | |
… | |
1376 | data. |
1438 | data. |
1377 | |
1439 | |
1378 | Example: read 2 bytes. |
1440 | Example: read 2 bytes. |
1379 | |
1441 | |
1380 | $handle->push_read (chunk => 2, sub { |
1442 | $handle->push_read (chunk => 2, sub { |
1381 | warn "yay ", unpack "H*", $_[1]; |
1443 | say "yay " . unpack "H*", $_[1]; |
1382 | }); |
1444 | }); |
1383 | |
1445 | |
1384 | =cut |
1446 | =cut |
1385 | |
1447 | |
1386 | register_read_type chunk => sub { |
1448 | register_read_type chunk => sub { |
… | |
… | |
1420 | if (@_ < 3) { |
1482 | if (@_ < 3) { |
1421 | # this is more than twice as fast as the generic code below |
1483 | # this is more than twice as fast as the generic code below |
1422 | sub { |
1484 | sub { |
1423 | $_[0]{rbuf} =~ s/^([^\015\012]*)(\015?\012)// or return; |
1485 | $_[0]{rbuf} =~ s/^([^\015\012]*)(\015?\012)// or return; |
1424 | |
1486 | |
1425 | $cb->($_[0], $1, $2); |
1487 | $cb->($_[0], "$1", "$2"); |
1426 | 1 |
1488 | 1 |
1427 | } |
1489 | } |
1428 | } else { |
1490 | } else { |
1429 | $eol = quotemeta $eol unless ref $eol; |
1491 | $eol = quotemeta $eol unless ref $eol; |
1430 | $eol = qr|^(.*?)($eol)|s; |
1492 | $eol = qr|^(.*?)($eol)|s; |
1431 | |
1493 | |
1432 | sub { |
1494 | sub { |
1433 | $_[0]{rbuf} =~ s/$eol// or return; |
1495 | $_[0]{rbuf} =~ s/$eol// or return; |
1434 | |
1496 | |
1435 | $cb->($_[0], $1, $2); |
1497 | $cb->($_[0], "$1", "$2"); |
1436 | 1 |
1498 | 1 |
1437 | } |
1499 | } |
1438 | } |
1500 | } |
1439 | }; |
1501 | }; |
1440 | |
1502 | |
… | |
… | |
1488 | |
1550 | |
1489 | sub { |
1551 | sub { |
1490 | # accept |
1552 | # accept |
1491 | if ($$rbuf =~ $accept) { |
1553 | if ($$rbuf =~ $accept) { |
1492 | $data .= substr $$rbuf, 0, $+[0], ""; |
1554 | $data .= substr $$rbuf, 0, $+[0], ""; |
1493 | $cb->($self, $data); |
1555 | $cb->($_[0], $data); |
1494 | return 1; |
1556 | return 1; |
1495 | } |
1557 | } |
1496 | |
1558 | |
1497 | # reject |
1559 | # reject |
1498 | if ($reject && $$rbuf =~ $reject) { |
1560 | if ($reject && $$rbuf =~ $reject) { |
1499 | $self->_error (Errno::EBADMSG); |
1561 | $_[0]->_error (Errno::EBADMSG); |
1500 | } |
1562 | } |
1501 | |
1563 | |
1502 | # skip |
1564 | # skip |
1503 | if ($skip && $$rbuf =~ $skip) { |
1565 | if ($skip && $$rbuf =~ $skip) { |
1504 | $data .= substr $$rbuf, 0, $+[0], ""; |
1566 | $data .= substr $$rbuf, 0, $+[0], ""; |
… | |
… | |
1520 | my ($self, $cb) = @_; |
1582 | my ($self, $cb) = @_; |
1521 | |
1583 | |
1522 | sub { |
1584 | sub { |
1523 | unless ($_[0]{rbuf} =~ s/^(0|[1-9][0-9]*)://) { |
1585 | unless ($_[0]{rbuf} =~ s/^(0|[1-9][0-9]*)://) { |
1524 | if ($_[0]{rbuf} =~ /[^0-9]/) { |
1586 | if ($_[0]{rbuf} =~ /[^0-9]/) { |
1525 | $self->_error (Errno::EBADMSG); |
1587 | $_[0]->_error (Errno::EBADMSG); |
1526 | } |
1588 | } |
1527 | return; |
1589 | return; |
1528 | } |
1590 | } |
1529 | |
1591 | |
1530 | my $len = $1; |
1592 | my $len = $1; |
1531 | |
1593 | |
1532 | $self->unshift_read (chunk => $len, sub { |
1594 | $_[0]->unshift_read (chunk => $len, sub { |
1533 | my $string = $_[1]; |
1595 | my $string = $_[1]; |
1534 | $_[0]->unshift_read (chunk => 1, sub { |
1596 | $_[0]->unshift_read (chunk => 1, sub { |
1535 | if ($_[1] eq ",") { |
1597 | if ($_[1] eq ",") { |
1536 | $cb->($_[0], $string); |
1598 | $cb->($_[0], $string); |
1537 | } else { |
1599 | } else { |
1538 | $self->_error (Errno::EBADMSG); |
1600 | $_[0]->_error (Errno::EBADMSG); |
1539 | } |
1601 | } |
1540 | }); |
1602 | }); |
1541 | }); |
1603 | }); |
1542 | |
1604 | |
1543 | 1 |
1605 | 1 |
… | |
… | |
1616 | |
1678 | |
1617 | my $data; |
1679 | my $data; |
1618 | my $rbuf = \$self->{rbuf}; |
1680 | my $rbuf = \$self->{rbuf}; |
1619 | |
1681 | |
1620 | sub { |
1682 | sub { |
1621 | my $ref = eval { $json->incr_parse ($self->{rbuf}) }; |
1683 | my $ref = eval { $json->incr_parse ($_[0]{rbuf}) }; |
1622 | |
1684 | |
1623 | if ($ref) { |
1685 | if ($ref) { |
1624 | $self->{rbuf} = $json->incr_text; |
1686 | $_[0]{rbuf} = $json->incr_text; |
1625 | $json->incr_text = ""; |
1687 | $json->incr_text = ""; |
1626 | $cb->($self, $ref); |
1688 | $cb->($_[0], $ref); |
1627 | |
1689 | |
1628 | 1 |
1690 | 1 |
1629 | } elsif ($@) { |
1691 | } elsif ($@) { |
1630 | # error case |
1692 | # error case |
1631 | $json->incr_skip; |
1693 | $json->incr_skip; |
1632 | |
1694 | |
1633 | $self->{rbuf} = $json->incr_text; |
1695 | $_[0]{rbuf} = $json->incr_text; |
1634 | $json->incr_text = ""; |
1696 | $json->incr_text = ""; |
1635 | |
1697 | |
1636 | $self->_error (Errno::EBADMSG); |
1698 | $_[0]->_error (Errno::EBADMSG); |
1637 | |
1699 | |
1638 | () |
1700 | () |
1639 | } else { |
1701 | } else { |
1640 | $self->{rbuf} = ""; |
1702 | $_[0]{rbuf} = ""; |
1641 | |
1703 | |
1642 | () |
1704 | () |
1643 | } |
1705 | } |
1644 | } |
1706 | } |
1645 | }; |
1707 | }; |
… | |
… | |
1655 | =cut |
1717 | =cut |
1656 | |
1718 | |
1657 | register_read_type storable => sub { |
1719 | register_read_type storable => sub { |
1658 | my ($self, $cb) = @_; |
1720 | my ($self, $cb) = @_; |
1659 | |
1721 | |
1660 | require Storable; |
1722 | require Storable unless $Storable::VERSION; |
1661 | |
1723 | |
1662 | sub { |
1724 | sub { |
1663 | # when we can use 5.10 we can use ".", but for 5.8 we use the re-pack method |
1725 | # when we can use 5.10 we can use ".", but for 5.8 we use the re-pack method |
1664 | defined (my $len = eval { unpack "w", $_[0]{rbuf} }) |
1726 | defined (my $len = eval { unpack "w", $_[0]{rbuf} }) |
1665 | or return; |
1727 | or return; |
… | |
… | |
1678 | # read remaining chunk |
1740 | # read remaining chunk |
1679 | $_[0]->unshift_read (chunk => $len, sub { |
1741 | $_[0]->unshift_read (chunk => $len, sub { |
1680 | if (my $ref = eval { Storable::thaw ($_[1]) }) { |
1742 | if (my $ref = eval { Storable::thaw ($_[1]) }) { |
1681 | $cb->($_[0], $ref); |
1743 | $cb->($_[0], $ref); |
1682 | } else { |
1744 | } else { |
1683 | $self->_error (Errno::EBADMSG); |
1745 | $_[0]->_error (Errno::EBADMSG); |
1684 | } |
1746 | } |
1685 | }); |
1747 | }); |
1686 | } |
1748 | } |
1687 | |
1749 | |
1688 | 1 |
1750 | 1 |
… | |
… | |
1726 | Note that AnyEvent::Handle will automatically C<start_read> for you when |
1788 | Note that AnyEvent::Handle will automatically C<start_read> for you when |
1727 | you change the C<on_read> callback or push/unshift a read callback, and it |
1789 | you change the C<on_read> callback or push/unshift a read callback, and it |
1728 | will automatically C<stop_read> for you when neither C<on_read> is set nor |
1790 | will automatically C<stop_read> for you when neither C<on_read> is set nor |
1729 | there are any read requests in the queue. |
1791 | there are any read requests in the queue. |
1730 | |
1792 | |
1731 | These methods will have no effect when in TLS mode (as TLS doesn't support |
1793 | In older versions of this module (<= 5.3), these methods had no effect, |
1732 | half-duplex connections). |
1794 | as TLS does not support half-duplex connections. In current versions they |
|
|
1795 | work as expected, as this behaviour is required to avoid certain resource |
|
|
1796 | attacks, where the program would be forced to read (and buffer) arbitrary |
|
|
1797 | amounts of data before being able to send some data. The drawback is that |
|
|
1798 | some readings of the the SSL/TLS specifications basically require this |
|
|
1799 | attack to be working, as SSL/TLS implementations might stall sending data |
|
|
1800 | during a rehandshake. |
|
|
1801 | |
|
|
1802 | As a guideline, during the initial handshake, you should not stop reading, |
|
|
1803 | and as a client, it might cause problems, depending on your application. |
1733 | |
1804 | |
1734 | =cut |
1805 | =cut |
1735 | |
1806 | |
1736 | sub stop_read { |
1807 | sub stop_read { |
1737 | my ($self) = @_; |
1808 | my ($self) = @_; |
1738 | |
1809 | |
1739 | delete $self->{_rw} unless $self->{tls}; |
1810 | delete $self->{_rw}; |
1740 | } |
1811 | } |
1741 | |
1812 | |
1742 | sub start_read { |
1813 | sub start_read { |
1743 | my ($self) = @_; |
1814 | my ($self) = @_; |
1744 | |
1815 | |
1745 | unless ($self->{_rw} || $self->{_eof} || !$self->{fh}) { |
1816 | unless ($self->{_rw} || $self->{_eof} || !$self->{fh}) { |
1746 | Scalar::Util::weaken $self; |
1817 | Scalar::Util::weaken $self; |
1747 | |
1818 | |
1748 | $self->{_rw} = AE::io $self->{fh}, 0, sub { |
1819 | $self->{_rw} = AE::io $self->{fh}, 0, sub { |
1749 | my $rbuf = \($self->{tls} ? my $buf : $self->{rbuf}); |
1820 | my $rbuf = \($self->{tls} ? my $buf : $self->{rbuf}); |
1750 | my $len = sysread $self->{fh}, $$rbuf, $self->{read_size} || 8192, length $$rbuf; |
1821 | my $len = sysread $self->{fh}, $$rbuf, $self->{read_size}, length $$rbuf; |
1751 | |
1822 | |
1752 | if ($len > 0) { |
1823 | if ($len > 0) { |
1753 | $self->{_activity} = $self->{_ractivity} = AE::now; |
1824 | $self->{_activity} = $self->{_ractivity} = AE::now; |
1754 | |
1825 | |
1755 | if ($self->{tls}) { |
1826 | if ($self->{tls}) { |
1756 | Net::SSLeay::BIO_write ($self->{_rbio}, $$rbuf); |
1827 | Net::SSLeay::BIO_write ($self->{_rbio}, $$rbuf); |
1757 | |
1828 | |
1758 | &_dotls ($self); |
1829 | &_dotls ($self); |
1759 | } else { |
1830 | } else { |
1760 | $self->_drain_rbuf; |
1831 | $self->_drain_rbuf; |
|
|
1832 | } |
|
|
1833 | |
|
|
1834 | if ($len == $self->{read_size}) { |
|
|
1835 | $self->{read_size} *= 2; |
|
|
1836 | $self->{read_size} = $self->{max_read_size} || MAX_READ_SIZE |
|
|
1837 | if $self->{read_size} > ($self->{max_read_size} || MAX_READ_SIZE); |
1761 | } |
1838 | } |
1762 | |
1839 | |
1763 | } elsif (defined $len) { |
1840 | } elsif (defined $len) { |
1764 | delete $self->{_rw}; |
1841 | delete $self->{_rw}; |
1765 | $self->{_eof} = 1; |
1842 | $self->{_eof} = 1; |
… | |
… | |
1940 | Net::SSLeay::CTX_set_mode ($tls, 1|2); |
2017 | Net::SSLeay::CTX_set_mode ($tls, 1|2); |
1941 | |
2018 | |
1942 | $self->{_rbio} = Net::SSLeay::BIO_new (Net::SSLeay::BIO_s_mem ()); |
2019 | $self->{_rbio} = Net::SSLeay::BIO_new (Net::SSLeay::BIO_s_mem ()); |
1943 | $self->{_wbio} = Net::SSLeay::BIO_new (Net::SSLeay::BIO_s_mem ()); |
2020 | $self->{_wbio} = Net::SSLeay::BIO_new (Net::SSLeay::BIO_s_mem ()); |
1944 | |
2021 | |
1945 | Net::SSLeay::BIO_write ($self->{_rbio}, delete $self->{rbuf}); |
2022 | Net::SSLeay::BIO_write ($self->{_rbio}, $self->{rbuf}); |
|
|
2023 | $self->{rbuf} = ""; |
1946 | |
2024 | |
1947 | Net::SSLeay::set_bio ($tls, $self->{_rbio}, $self->{_wbio}); |
2025 | Net::SSLeay::set_bio ($tls, $self->{_rbio}, $self->{_wbio}); |
1948 | |
2026 | |
1949 | $self->{_on_starttls} = sub { $_[0]{on_starttls}(@_) } |
2027 | $self->{_on_starttls} = sub { $_[0]{on_starttls}(@_) } |
1950 | if $self->{on_starttls}; |
2028 | if $self->{on_starttls}; |
… | |
… | |
1987 | $self->{tls_ctx}->_put_session (delete $self->{tls}) |
2065 | $self->{tls_ctx}->_put_session (delete $self->{tls}) |
1988 | if $self->{tls} > 0; |
2066 | if $self->{tls} > 0; |
1989 | |
2067 | |
1990 | delete @$self{qw(_rbio _wbio _tls_wbuf _on_starttls)}; |
2068 | delete @$self{qw(_rbio _wbio _tls_wbuf _on_starttls)}; |
1991 | } |
2069 | } |
|
|
2070 | |
|
|
2071 | =item $handle->resettls |
|
|
2072 | |
|
|
2073 | This rarely-used method simply resets and TLS state on the handle, usually |
|
|
2074 | causing data loss. |
|
|
2075 | |
|
|
2076 | One case where it may be useful is when you want to skip over the data in |
|
|
2077 | the stream but you are not interested in interpreting it, so data loss is |
|
|
2078 | no concern. |
|
|
2079 | |
|
|
2080 | =cut |
|
|
2081 | |
|
|
2082 | *resettls = \&_freetls; |
1992 | |
2083 | |
1993 | sub DESTROY { |
2084 | sub DESTROY { |
1994 | my ($self) = @_; |
2085 | my ($self) = @_; |
1995 | |
2086 | |
1996 | &_freetls; |
2087 | &_freetls; |
… | |
… | |
2112 | |
2203 | |
2113 | It is only safe to "forget" the reference inside EOF or error callbacks, |
2204 | It is only safe to "forget" the reference inside EOF or error callbacks, |
2114 | from within all other callbacks, you need to explicitly call the C<< |
2205 | from within all other callbacks, you need to explicitly call the C<< |
2115 | ->destroy >> method. |
2206 | ->destroy >> method. |
2116 | |
2207 | |
|
|
2208 | =item Why is my C<on_eof> callback never called? |
|
|
2209 | |
|
|
2210 | Probably because your C<on_error> callback is being called instead: When |
|
|
2211 | you have outstanding requests in your read queue, then an EOF is |
|
|
2212 | considered an error as you clearly expected some data. |
|
|
2213 | |
|
|
2214 | To avoid this, make sure you have an empty read queue whenever your handle |
|
|
2215 | is supposed to be "idle" (i.e. connection closes are O.K.). You can set |
|
|
2216 | an C<on_read> handler that simply pushes the first read requests in the |
|
|
2217 | queue. |
|
|
2218 | |
|
|
2219 | See also the next question, which explains this in a bit more detail. |
|
|
2220 | |
|
|
2221 | =item How can I serve requests in a loop? |
|
|
2222 | |
|
|
2223 | Most protocols consist of some setup phase (authentication for example) |
|
|
2224 | followed by a request handling phase, where the server waits for requests |
|
|
2225 | and handles them, in a loop. |
|
|
2226 | |
|
|
2227 | There are two important variants: The first (traditional, better) variant |
|
|
2228 | handles requests until the server gets some QUIT command, causing it to |
|
|
2229 | close the connection first (highly desirable for a busy TCP server). A |
|
|
2230 | client dropping the connection is an error, which means this variant can |
|
|
2231 | detect an unexpected detection close. |
|
|
2232 | |
|
|
2233 | To handle this case, always make sure you have a on-empty read queue, by |
|
|
2234 | pushing the "read request start" handler on it: |
|
|
2235 | |
|
|
2236 | # we assume a request starts with a single line |
|
|
2237 | my @start_request; @start_request = (line => sub { |
|
|
2238 | my ($hdl, $line) = @_; |
|
|
2239 | |
|
|
2240 | ... handle request |
|
|
2241 | |
|
|
2242 | # push next request read, possibly from a nested callback |
|
|
2243 | $hdl->push_read (@start_request); |
|
|
2244 | }); |
|
|
2245 | |
|
|
2246 | # auth done, now go into request handling loop |
|
|
2247 | # now push the first @start_request |
|
|
2248 | $hdl->push_read (@start_request); |
|
|
2249 | |
|
|
2250 | By always having an outstanding C<push_read>, the handle always expects |
|
|
2251 | some data and raises the C<EPIPE> error when the connction is dropped |
|
|
2252 | unexpectedly. |
|
|
2253 | |
|
|
2254 | The second variant is a protocol where the client can drop the connection |
|
|
2255 | at any time. For TCP, this means that the server machine may run out of |
|
|
2256 | sockets easier, and in general, it means you cannot distinguish a protocl |
|
|
2257 | failure/client crash from a normal connection close. Nevertheless, these |
|
|
2258 | kinds of protocols are common (and sometimes even the best solution to the |
|
|
2259 | problem). |
|
|
2260 | |
|
|
2261 | Having an outstanding read request at all times is possible if you ignore |
|
|
2262 | C<EPIPE> errors, but this doesn't help with when the client drops the |
|
|
2263 | connection during a request, which would still be an error. |
|
|
2264 | |
|
|
2265 | A better solution is to push the initial request read in an C<on_read> |
|
|
2266 | callback. This avoids an error, as when the server doesn't expect data |
|
|
2267 | (i.e. is idly waiting for the next request, an EOF will not raise an |
|
|
2268 | error, but simply result in an C<on_eof> callback. It is also a bit slower |
|
|
2269 | and simpler: |
|
|
2270 | |
|
|
2271 | # auth done, now go into request handling loop |
|
|
2272 | $hdl->on_read (sub { |
|
|
2273 | my ($hdl) = @_; |
|
|
2274 | |
|
|
2275 | # called each time we receive data but the read queue is empty |
|
|
2276 | # simply start read the request |
|
|
2277 | |
|
|
2278 | $hdl->push_read (line => sub { |
|
|
2279 | my ($hdl, $line) = @_; |
|
|
2280 | |
|
|
2281 | ... handle request |
|
|
2282 | |
|
|
2283 | # do nothing special when the request has been handled, just |
|
|
2284 | # let the request queue go empty. |
|
|
2285 | }); |
|
|
2286 | }); |
|
|
2287 | |
2117 | =item I get different callback invocations in TLS mode/Why can't I pause |
2288 | =item I get different callback invocations in TLS mode/Why can't I pause |
2118 | reading? |
2289 | reading? |
2119 | |
2290 | |
2120 | Unlike, say, TCP, TLS connections do not consist of two independent |
2291 | Unlike, say, TCP, TLS connections do not consist of two independent |
2121 | communication channels, one for each direction. Or put differently, the |
2292 | communication channels, one for each direction. Or put differently, the |
… | |
… | |
2142 | $handle->on_eof (undef); |
2313 | $handle->on_eof (undef); |
2143 | $handle->on_error (sub { |
2314 | $handle->on_error (sub { |
2144 | my $data = delete $_[0]{rbuf}; |
2315 | my $data = delete $_[0]{rbuf}; |
2145 | }); |
2316 | }); |
2146 | |
2317 | |
|
|
2318 | Note that this example removes the C<rbuf> member from the handle object, |
|
|
2319 | which is not normally allowed by the API. It is expressly permitted in |
|
|
2320 | this case only, as the handle object needs to be destroyed afterwards. |
|
|
2321 | |
2147 | The reason to use C<on_error> is that TCP connections, due to latencies |
2322 | The reason to use C<on_error> is that TCP connections, due to latencies |
2148 | and packets loss, might get closed quite violently with an error, when in |
2323 | and packets loss, might get closed quite violently with an error, when in |
2149 | fact all data has been received. |
2324 | fact all data has been received. |
2150 | |
2325 | |
2151 | It is usually better to use acknowledgements when transferring data, |
2326 | It is usually better to use acknowledgements when transferring data, |
… | |
… | |
2161 | C<low_water_mark> this will be called precisely when all data has been |
2336 | C<low_water_mark> this will be called precisely when all data has been |
2162 | written to the socket: |
2337 | written to the socket: |
2163 | |
2338 | |
2164 | $handle->push_write (...); |
2339 | $handle->push_write (...); |
2165 | $handle->on_drain (sub { |
2340 | $handle->on_drain (sub { |
2166 | warn "all data submitted to the kernel\n"; |
2341 | AE::log debug => "all data submitted to the kernel\n"; |
2167 | undef $handle; |
2342 | undef $handle; |
2168 | }); |
2343 | }); |
2169 | |
2344 | |
2170 | If you just want to queue some data and then signal EOF to the other side, |
2345 | If you just want to queue some data and then signal EOF to the other side, |
2171 | consider using C<< ->push_shutdown >> instead. |
2346 | consider using C<< ->push_shutdown >> instead. |