1 |
root |
1.1 |
=head1 NAME |
2 |
|
|
|
3 |
|
|
Coro::Channel - message queues |
4 |
|
|
|
5 |
|
|
=head1 SYNOPSIS |
6 |
|
|
|
7 |
|
|
use Coro::Channel; |
8 |
|
|
|
9 |
|
|
$q1 = new Coro::Channel <maxsize>; |
10 |
|
|
|
11 |
root |
1.46 |
$q1->put ("xxx"); |
12 |
root |
1.1 |
print $q1->get; |
13 |
|
|
|
14 |
|
|
die unless $q1->size; |
15 |
|
|
|
16 |
|
|
=head1 DESCRIPTION |
17 |
|
|
|
18 |
root |
1.75 |
A Coro::Channel is the equivalent of a unix pipe (and similar to amiga |
19 |
|
|
message ports): you can put things into it on one end and read things out |
20 |
|
|
of it from the other end. If the capacity of the Channel is maxed out |
21 |
|
|
writers will block. Both ends of a Channel can be read/written from by as |
22 |
|
|
many coroutines as you want concurrently. |
23 |
root |
1.8 |
|
24 |
root |
1.1 |
=over 4 |
25 |
|
|
|
26 |
|
|
=cut |
27 |
|
|
|
28 |
|
|
package Coro::Channel; |
29 |
|
|
|
30 |
root |
1.92 |
use common::sense; |
31 |
root |
1.45 |
|
32 |
root |
1.3 |
use Coro (); |
33 |
root |
1.71 |
use Coro::Semaphore (); |
34 |
root |
1.1 |
|
35 |
root |
1.99 |
our $VERSION = 5.24; |
36 |
root |
1.1 |
|
37 |
root |
1.70 |
sub DATA (){ 0 } |
38 |
root |
1.71 |
sub SGET (){ 1 } |
39 |
|
|
sub SPUT (){ 2 } |
40 |
root |
1.70 |
|
41 |
root |
1.1 |
=item $q = new Coro:Channel $maxsize |
42 |
|
|
|
43 |
root |
1.75 |
Create a new channel with the given maximum size (practically unlimited |
44 |
|
|
if C<maxsize> is omitted). Giving a size of one gives you a traditional |
45 |
|
|
channel, i.e. a queue that can store only a single element (which means |
46 |
|
|
there will be no buffering, and C<put> will wait until there is a |
47 |
|
|
corresponding C<get> call). To buffer one element you have to specify |
48 |
|
|
C<2>, and so on. |
49 |
root |
1.1 |
|
50 |
|
|
=cut |
51 |
|
|
|
52 |
|
|
sub new { |
53 |
root |
1.84 |
# we cheat and set infinity == 2*10**9 |
54 |
root |
1.71 |
bless [ |
55 |
root |
1.96 |
[], # initially empty |
56 |
|
|
(Coro::Semaphore::_alloc 0), # counts data |
57 |
|
|
(Coro::Semaphore::_alloc +($_[1] || 2_000_000_000) - 1), # counts remaining space |
58 |
root |
1.71 |
] |
59 |
root |
1.1 |
} |
60 |
|
|
|
61 |
root |
1.44 |
=item $q->put ($scalar) |
62 |
root |
1.1 |
|
63 |
|
|
Put the given scalar into the queue. |
64 |
|
|
|
65 |
|
|
=cut |
66 |
|
|
|
67 |
|
|
sub put { |
68 |
root |
1.70 |
push @{$_[0][DATA]}, $_[1]; |
69 |
root |
1.71 |
Coro::Semaphore::up $_[0][SGET]; |
70 |
|
|
Coro::Semaphore::down $_[0][SPUT]; |
71 |
root |
1.1 |
} |
72 |
|
|
|
73 |
|
|
=item $q->get |
74 |
|
|
|
75 |
|
|
Return the next element from the queue, waiting if necessary. |
76 |
|
|
|
77 |
|
|
=cut |
78 |
|
|
|
79 |
|
|
sub get { |
80 |
root |
1.71 |
Coro::Semaphore::down $_[0][SGET]; |
81 |
|
|
Coro::Semaphore::up $_[0][SPUT]; |
82 |
root |
1.70 |
shift @{$_[0][DATA]} |
83 |
root |
1.47 |
} |
84 |
|
|
|
85 |
root |
1.77 |
=item $q->shutdown |
86 |
|
|
|
87 |
|
|
Shuts down the Channel by pushing a virtual end marker onto it: This |
88 |
|
|
changes the behaviour of the Channel when it becomes or is empty to return |
89 |
|
|
C<undef>, almost as if infinitely many C<undef> elements have been put |
90 |
|
|
into the queue. |
91 |
|
|
|
92 |
|
|
Specifically, this function wakes up any pending C<get> calls and lets |
93 |
|
|
them return C<undef>, the same on future C<get> calls. C<size> will return |
94 |
|
|
the real number of stored elements, though. |
95 |
|
|
|
96 |
|
|
Another way to describe the behaviour is that C<get> calls will not block |
97 |
|
|
when the queue becomes empty but immediately return C<undef>. This means |
98 |
|
|
that calls to C<put> will work normally and the data will be returned on |
99 |
|
|
subsequent C<get> calls. |
100 |
|
|
|
101 |
|
|
This method is useful to signal the end of data to any consumers, quite |
102 |
|
|
similar to an end of stream on e.g. a tcp socket: You have one or more |
103 |
|
|
producers that C<put> data into the Channel and one or more consumers who |
104 |
|
|
C<get> them. When all producers have finished producing data, a call to |
105 |
|
|
C<shutdown> signals this fact to any consumers. |
106 |
|
|
|
107 |
|
|
=cut |
108 |
|
|
|
109 |
|
|
sub shutdown { |
110 |
root |
1.79 |
Coro::Semaphore::adjust $_[0][SGET], 2_000_000_000; |
111 |
root |
1.77 |
} |
112 |
|
|
|
113 |
root |
1.1 |
=item $q->size |
114 |
|
|
|
115 |
|
|
Return the number of elements waiting to be consumed. Please note that: |
116 |
|
|
|
117 |
|
|
if ($q->size) { |
118 |
|
|
my $data = $q->get; |
119 |
root |
1.77 |
... |
120 |
root |
1.1 |
} |
121 |
|
|
|
122 |
root |
1.83 |
is I<not> a race condition but instead works just fine. Note that the |
123 |
|
|
number of elements that wait can be larger than C<$maxsize>, as it |
124 |
|
|
includes any coroutines waiting to put data into the channel (but not any |
125 |
|
|
shutdown condition). |
126 |
|
|
|
127 |
root |
1.96 |
This means that the number returned is I<precisely> the number of calls |
128 |
|
|
to C<get> that will succeed instantly and return some data. Calling |
129 |
|
|
C<shutdown> has no effect on this number. |
130 |
root |
1.1 |
|
131 |
|
|
=cut |
132 |
|
|
|
133 |
|
|
sub size { |
134 |
root |
1.70 |
scalar @{$_[0][DATA]} |
135 |
root |
1.1 |
} |
136 |
|
|
|
137 |
root |
1.96 |
# this is not undocumented by accident - if it breaks, you |
138 |
|
|
# get to keep the pieces |
139 |
|
|
sub adjust { |
140 |
|
|
Coro::Semaphore::adjust $_[0][SPUT], $_[1]; |
141 |
|
|
} |
142 |
|
|
|
143 |
root |
1.1 |
1; |
144 |
|
|
|
145 |
|
|
=back |
146 |
|
|
|
147 |
|
|
=head1 AUTHOR |
148 |
|
|
|
149 |
root |
1.35 |
Marc Lehmann <schmorp@schmorp.de> |
150 |
root |
1.33 |
http://home.schmorp.de/ |
151 |
root |
1.1 |
|
152 |
|
|
=cut |
153 |
|
|
|