line |
stmt |
bran |
cond |
sub |
pod |
time |
code |
1
|
|
|
|
|
|
|
# You may distribute under the terms of either the GNU General Public License |
2
|
|
|
|
|
|
|
# or the Artistic License (the same terms as Perl itself) |
3
|
|
|
|
|
|
|
# |
4
|
|
|
|
|
|
|
# (C) Paul Evans, 2010 -- leonerd@leonerd.org.uk |
5
|
|
|
|
|
|
|
|
6
|
|
|
|
|
|
|
package CPS::Functional; |
7
|
|
|
|
|
|
|
|
8
|
5
|
|
|
5
|
|
110646
|
use strict; |
|
5
|
|
|
|
|
13
|
|
|
5
|
|
|
|
|
183
|
|
9
|
5
|
|
|
5
|
|
28
|
use warnings; |
|
5
|
|
|
|
|
10
|
|
|
5
|
|
|
|
|
222
|
|
10
|
|
|
|
|
|
|
|
11
|
|
|
|
|
|
|
our $VERSION = '0.18'; |
12
|
|
|
|
|
|
|
|
13
|
5
|
|
|
5
|
|
27
|
use Carp; |
|
5
|
|
|
|
|
9
|
|
|
5
|
|
|
|
|
465
|
|
14
|
|
|
|
|
|
|
|
15
|
5
|
|
|
5
|
|
24
|
use Exporter 'import'; |
|
5
|
|
|
|
|
7
|
|
|
5
|
|
|
|
|
167
|
|
16
|
|
|
|
|
|
|
|
17
|
5
|
|
|
5
|
|
2387
|
use CPS qw( gkloop ); |
|
5
|
|
|
|
|
15
|
|
|
5
|
|
|
|
|
845
|
|
18
|
|
|
|
|
|
|
|
19
|
|
|
|
|
|
|
our @CPS_PRIMS = qw( |
20
|
|
|
|
|
|
|
kmap |
21
|
|
|
|
|
|
|
kgrep |
22
|
|
|
|
|
|
|
kfoldl kfoldr |
23
|
|
|
|
|
|
|
kunfold |
24
|
|
|
|
|
|
|
); |
25
|
|
|
|
|
|
|
|
26
|
|
|
|
|
|
|
our @EXPORT_OK = ( |
27
|
|
|
|
|
|
|
@CPS_PRIMS, |
28
|
|
|
|
|
|
|
map( "g$_", @CPS_PRIMS ), |
29
|
|
|
|
|
|
|
); |
30
|
|
|
|
|
|
|
|
31
|
|
|
|
|
|
|
# Don't hard-depend on Sub::Name since it's only a niceness for stack traces |
32
|
|
|
|
|
|
|
BEGIN { |
33
|
5
|
50
|
|
5
|
|
9
|
if( eval { require Sub::Name } ) { |
|
5
|
|
|
|
|
41
|
|
34
|
5
|
|
|
|
|
6872
|
*subname = \&Sub::Name::subname; |
35
|
|
|
|
|
|
|
} |
36
|
|
|
|
|
|
|
else { |
37
|
|
|
|
|
|
|
# Ignore the name, return the CODEref |
38
|
0
|
|
|
|
|
0
|
*subname = sub { return $_[1] }; |
|
0
|
|
|
|
|
0
|
|
39
|
|
|
|
|
|
|
} |
40
|
|
|
|
|
|
|
} |
41
|
|
|
|
|
|
|
|
42
|
|
|
|
|
|
|
=head1 NAME |
43
|
|
|
|
|
|
|
|
44
|
|
|
|
|
|
|
C - functional utilities in Continuation-Passing Style |
45
|
|
|
|
|
|
|
|
46
|
|
|
|
|
|
|
=head1 SYNOPSIS |
47
|
|
|
|
|
|
|
|
48
|
|
|
|
|
|
|
use CPS::Functional qw( kmap ); |
49
|
|
|
|
|
|
|
|
50
|
|
|
|
|
|
|
use Example::HTTP::Client qw( k_get_http ); |
51
|
|
|
|
|
|
|
use List::Util qw( sum ); |
52
|
|
|
|
|
|
|
|
53
|
|
|
|
|
|
|
my @URLs = ( |
54
|
|
|
|
|
|
|
"http://www.foo.com", |
55
|
|
|
|
|
|
|
"http://www.bar.com", |
56
|
|
|
|
|
|
|
); |
57
|
|
|
|
|
|
|
|
58
|
|
|
|
|
|
|
kmap( \@URLs, |
59
|
|
|
|
|
|
|
sub { |
60
|
|
|
|
|
|
|
my ( $item, $kret ) = @_; |
61
|
|
|
|
|
|
|
|
62
|
|
|
|
|
|
|
k_get_http( uri => $item, on_response => sub { |
63
|
|
|
|
|
|
|
my ( $response ) = @_; |
64
|
|
|
|
|
|
|
|
65
|
|
|
|
|
|
|
$kret->( $response->content_length ); |
66
|
|
|
|
|
|
|
} ); |
67
|
|
|
|
|
|
|
}, |
68
|
|
|
|
|
|
|
sub { |
69
|
|
|
|
|
|
|
my ( @sizes ) = @_; |
70
|
|
|
|
|
|
|
|
71
|
|
|
|
|
|
|
say "Total length of all URLs: " . sum(@sizes); |
72
|
|
|
|
|
|
|
}, |
73
|
|
|
|
|
|
|
); |
74
|
|
|
|
|
|
|
|
75
|
|
|
|
|
|
|
=head1 DESCRIPTION |
76
|
|
|
|
|
|
|
|
77
|
|
|
|
|
|
|
This module provides L versions of data-flow functionals, such as Perl's |
78
|
|
|
|
|
|
|
C |
79
|
|
|
|
|
|
|
data, which the functional manages. They are built on top of the control-flow |
80
|
|
|
|
|
|
|
functionals provided by the C module itself. |
81
|
|
|
|
|
|
|
|
82
|
|
|
|
|
|
|
=cut |
83
|
|
|
|
|
|
|
|
84
|
|
|
|
|
|
|
=head1 FUNCTIONS |
85
|
|
|
|
|
|
|
|
86
|
|
|
|
|
|
|
=cut |
87
|
|
|
|
|
|
|
|
88
|
|
|
|
|
|
|
=head2 kmap( \@items, \&body, $k ) |
89
|
|
|
|
|
|
|
|
90
|
|
|
|
|
|
|
CPS version of perl's C |
91
|
|
|
|
|
|
|
element in C<@items>, capturing the list of values the body passes into its |
92
|
|
|
|
|
|
|
continuation. When the items are exhausted, C<$k> is invoked and passed a list |
93
|
|
|
|
|
|
|
of all the collected values. |
94
|
|
|
|
|
|
|
|
95
|
|
|
|
|
|
|
$body->( $item, $kret ) |
96
|
|
|
|
|
|
|
$kret->( @items_out ) |
97
|
|
|
|
|
|
|
|
98
|
|
|
|
|
|
|
$k->( @all_items_out ) |
99
|
|
|
|
|
|
|
|
100
|
|
|
|
|
|
|
=cut |
101
|
|
|
|
|
|
|
|
102
|
|
|
|
|
|
|
sub gkmap |
103
|
|
|
|
|
|
|
{ |
104
|
1
|
|
|
1
|
0
|
5
|
my ( $gov, $items, $body, $k ) = @_; |
105
|
|
|
|
|
|
|
|
106
|
1
|
50
|
|
|
|
8
|
ref $items eq "ARRAY" or croak 'Expected $items as ARRAY ref'; |
107
|
1
|
50
|
|
|
|
5
|
ref $body eq "CODE" or croak 'Expected $body as CODE ref'; |
108
|
|
|
|
|
|
|
|
109
|
1
|
|
|
|
|
2
|
my @ret; |
110
|
1
|
|
|
|
|
2
|
my $idx = 0; |
111
|
|
|
|
|
|
|
|
112
|
|
|
|
|
|
|
gkloop( $gov, |
113
|
|
|
|
|
|
|
sub { |
114
|
4
|
|
|
4
|
|
8
|
my ( $knext, $klast ) = @_; |
115
|
4
|
100
|
|
|
|
12
|
goto &$klast unless $idx < scalar @$items; |
116
|
|
|
|
|
|
|
@_ = ( |
117
|
|
|
|
|
|
|
$items->[$idx++], |
118
|
3
|
|
|
|
|
16
|
sub { push @ret, @_; goto &$knext } |
|
3
|
|
|
|
|
9
|
|
119
|
3
|
|
|
|
|
15
|
); |
120
|
3
|
|
|
|
|
10
|
goto &$body; |
121
|
|
|
|
|
|
|
}, |
122
|
1
|
|
|
1
|
|
4
|
sub { $k->( @ret ) }, |
123
|
1
|
|
|
|
|
13
|
); |
124
|
|
|
|
|
|
|
} |
125
|
|
|
|
|
|
|
|
126
|
|
|
|
|
|
|
=head2 kgrep( \@items, \&body, $k ) |
127
|
|
|
|
|
|
|
|
128
|
|
|
|
|
|
|
CPS version of perl's C statement. Calls the C code once for each |
129
|
|
|
|
|
|
|
element in C<@items>, capturing those elements where the body's continuation |
130
|
|
|
|
|
|
|
was invoked with a true value. When the items are exhausted, C<$k> is invoked |
131
|
|
|
|
|
|
|
and passed a list of the subset of C<@items> which were selected. |
132
|
|
|
|
|
|
|
|
133
|
|
|
|
|
|
|
$body->( $item, $kret ) |
134
|
|
|
|
|
|
|
$kret->( $select ) |
135
|
|
|
|
|
|
|
|
136
|
|
|
|
|
|
|
$k->( @chosen_items ) |
137
|
|
|
|
|
|
|
|
138
|
|
|
|
|
|
|
=cut |
139
|
|
|
|
|
|
|
|
140
|
|
|
|
|
|
|
sub gkgrep |
141
|
|
|
|
|
|
|
{ |
142
|
1
|
|
|
1
|
0
|
4
|
my ( $gov, $items, $body, $k ) = @_; |
143
|
|
|
|
|
|
|
|
144
|
1
|
50
|
|
|
|
8
|
ref $items eq "ARRAY" or croak 'Expected $items as ARRAY ref'; |
145
|
1
|
50
|
|
|
|
6
|
ref $body eq "CODE" or croak 'Expected $body as CODE ref'; |
146
|
|
|
|
|
|
|
|
147
|
1
|
|
|
|
|
2
|
my @ret; |
148
|
1
|
|
|
|
|
3
|
my $idx = 0; |
149
|
|
|
|
|
|
|
|
150
|
|
|
|
|
|
|
gkloop( $gov, |
151
|
|
|
|
|
|
|
sub { |
152
|
5
|
|
|
5
|
|
8
|
my ( $knext, $klast ) = @_; |
153
|
5
|
100
|
|
|
|
15
|
goto &$klast unless $idx < scalar @$items; |
154
|
4
|
|
|
|
|
8
|
my $item = $items->[$idx++]; |
155
|
|
|
|
|
|
|
@_ = ( |
156
|
|
|
|
|
|
|
$item, |
157
|
4
|
100
|
|
|
|
25
|
sub { push @ret, $item if $_[0]; goto &$knext } |
|
4
|
|
|
|
|
11
|
|
158
|
4
|
|
|
|
|
18
|
); |
159
|
4
|
|
|
|
|
12
|
goto &$body; |
160
|
|
|
|
|
|
|
}, |
161
|
1
|
|
|
1
|
|
5
|
sub { $k->( @ret ) }, |
162
|
1
|
|
|
|
|
14
|
); |
163
|
|
|
|
|
|
|
} |
164
|
|
|
|
|
|
|
|
165
|
|
|
|
|
|
|
=head2 kfoldl( \@items, \&body, $k ) |
166
|
|
|
|
|
|
|
|
167
|
|
|
|
|
|
|
CPS version of C, which collapses (or "folds") a list of |
168
|
|
|
|
|
|
|
values down to a single scalar, by successively accumulating values together. |
169
|
|
|
|
|
|
|
|
170
|
|
|
|
|
|
|
If C<@items> is empty, invokes C<$k> immediately, passing in C. |
171
|
|
|
|
|
|
|
|
172
|
|
|
|
|
|
|
If C<@items> contains a single value, invokes C<$k> immediately, passing in |
173
|
|
|
|
|
|
|
just that single value. |
174
|
|
|
|
|
|
|
|
175
|
|
|
|
|
|
|
Otherwise, initialises an accumulator variable with the first value in |
176
|
|
|
|
|
|
|
C<@items>, then for each additional item, invokes the C passing in the |
177
|
|
|
|
|
|
|
accumulator and the next item, storing back into the accumulator the value |
178
|
|
|
|
|
|
|
that C passed to its continuation. When the C<@items> are exhausted, it |
179
|
|
|
|
|
|
|
invokes C<$k>, passing in the final value of the accumulator. |
180
|
|
|
|
|
|
|
|
181
|
|
|
|
|
|
|
$body->( $acc, $item, $kret ) |
182
|
|
|
|
|
|
|
$kret->( $new_acc ) |
183
|
|
|
|
|
|
|
|
184
|
|
|
|
|
|
|
$k->( $final_acc ) |
185
|
|
|
|
|
|
|
|
186
|
|
|
|
|
|
|
Technically, this is not a true Scheme/Haskell-style C, as it does not |
187
|
|
|
|
|
|
|
take an initial value. (It is what Haskell calls C.) However, if such |
188
|
|
|
|
|
|
|
an initial value is required, this can be provided by |
189
|
|
|
|
|
|
|
|
190
|
|
|
|
|
|
|
kfoldl( [ $initial, @items ], \&body, $k ) |
191
|
|
|
|
|
|
|
|
192
|
|
|
|
|
|
|
=cut |
193
|
|
|
|
|
|
|
|
194
|
|
|
|
|
|
|
sub gkfoldl |
195
|
|
|
|
|
|
|
{ |
196
|
1
|
|
|
1
|
0
|
3
|
my ( $gov, $items, $body, $k ) = @_; |
197
|
|
|
|
|
|
|
|
198
|
1
|
50
|
|
|
|
9
|
ref $items eq "ARRAY" or croak 'Expected $items as ARRAY ref'; |
199
|
1
|
50
|
|
|
|
6
|
ref $body eq "CODE" or croak 'Expected $body as CODE ref'; |
200
|
|
|
|
|
|
|
|
201
|
1
|
50
|
|
|
|
5
|
$k->( undef ), return if @$items == 0; |
202
|
1
|
50
|
|
|
|
4
|
$k->( $items->[0] ), return if @$items == 1; |
203
|
|
|
|
|
|
|
|
204
|
1
|
|
|
|
|
2
|
my $idx = 0; |
205
|
1
|
|
|
|
|
3
|
my $acc = $items->[$idx++]; |
206
|
|
|
|
|
|
|
|
207
|
|
|
|
|
|
|
gkloop( $gov, |
208
|
|
|
|
|
|
|
sub { |
209
|
3
|
|
|
3
|
|
5
|
my ( $knext, $klast ) = @_; |
210
|
3
|
100
|
|
|
|
12
|
goto &$klast unless $idx < scalar @$items; |
211
|
|
|
|
|
|
|
@_ = ( |
212
|
|
|
|
|
|
|
$acc, |
213
|
|
|
|
|
|
|
$items->[$idx++], |
214
|
2
|
|
|
|
|
17
|
sub { $acc = shift; goto &$knext } |
|
2
|
|
|
|
|
6
|
|
215
|
2
|
|
|
|
|
11
|
); |
216
|
2
|
|
|
|
|
8
|
goto &$body; |
217
|
|
|
|
|
|
|
}, |
218
|
1
|
|
|
1
|
|
4
|
sub { $k->( $acc ) }, |
219
|
1
|
|
|
|
|
21
|
); |
220
|
|
|
|
|
|
|
} |
221
|
|
|
|
|
|
|
|
222
|
|
|
|
|
|
|
=head2 kfoldr( \@items, \&body, $k ) |
223
|
|
|
|
|
|
|
|
224
|
|
|
|
|
|
|
A right-associative version of C. Where C starts with the |
225
|
|
|
|
|
|
|
first two elements in C<@items> and works forward, C starts with the |
226
|
|
|
|
|
|
|
last two and works backward. |
227
|
|
|
|
|
|
|
|
228
|
|
|
|
|
|
|
$body->( $item, $acc, $kret ) |
229
|
|
|
|
|
|
|
$kret->( $new_acc ) |
230
|
|
|
|
|
|
|
|
231
|
|
|
|
|
|
|
$k->( $final_acc ) |
232
|
|
|
|
|
|
|
|
233
|
|
|
|
|
|
|
As before, an initial value can be provided by modifying the C<@items> array, |
234
|
|
|
|
|
|
|
though note it has to be last this time: |
235
|
|
|
|
|
|
|
|
236
|
|
|
|
|
|
|
kfoldr( [ @items, $initial ], \&body, $k ) |
237
|
|
|
|
|
|
|
|
238
|
|
|
|
|
|
|
=cut |
239
|
|
|
|
|
|
|
|
240
|
|
|
|
|
|
|
sub gkfoldr |
241
|
|
|
|
|
|
|
{ |
242
|
1
|
|
|
1
|
0
|
3
|
my ( $gov, $items, $body, $k ) = @_; |
243
|
|
|
|
|
|
|
|
244
|
1
|
50
|
|
|
|
5
|
ref $items eq "ARRAY" or croak 'Expected $items as ARRAY ref'; |
245
|
1
|
50
|
|
|
|
6
|
ref $body eq "CODE" or croak 'Expected $body as CODE ref'; |
246
|
|
|
|
|
|
|
|
247
|
1
|
50
|
|
|
|
5
|
$k->( undef ), return if @$items == 0; |
248
|
1
|
50
|
|
|
|
4
|
$k->( $items->[0] ), return if @$items == 1; |
249
|
|
|
|
|
|
|
|
250
|
1
|
|
|
|
|
4
|
my $idx = scalar(@$items) - 1; |
251
|
1
|
|
|
|
|
3
|
my $acc = $items->[$idx--]; |
252
|
|
|
|
|
|
|
|
253
|
|
|
|
|
|
|
gkloop( $gov, |
254
|
|
|
|
|
|
|
sub { |
255
|
3
|
|
|
3
|
|
5
|
my ( $knext, $klast ) = @_; |
256
|
3
|
100
|
|
|
|
10
|
goto &$klast if $idx < 0; |
257
|
|
|
|
|
|
|
@_ = ( |
258
|
|
|
|
|
|
|
$items->[$idx--], |
259
|
|
|
|
|
|
|
$acc, |
260
|
2
|
|
|
|
|
13
|
sub { $acc = shift; goto &$knext } |
|
2
|
|
|
|
|
5
|
|
261
|
2
|
|
|
|
|
10
|
); |
262
|
2
|
|
|
|
|
6
|
goto &$body; |
263
|
|
|
|
|
|
|
}, |
264
|
1
|
|
|
1
|
|
4
|
sub { $k->( $acc ) }, |
265
|
1
|
|
|
|
|
18
|
); |
266
|
|
|
|
|
|
|
} |
267
|
|
|
|
|
|
|
|
268
|
|
|
|
|
|
|
=head2 kunfold( $seed, \&body, $k ) |
269
|
|
|
|
|
|
|
|
270
|
|
|
|
|
|
|
An inverse operation to C; turns a single scalar into a list of |
271
|
|
|
|
|
|
|
items. Repeatedly calls the C code, capturing the values it returns, |
272
|
|
|
|
|
|
|
until it indicates the end of the loop, then invoke C<$k> with the collected |
273
|
|
|
|
|
|
|
values. |
274
|
|
|
|
|
|
|
|
275
|
|
|
|
|
|
|
$body->( $seed, $kmore, $kdone ) |
276
|
|
|
|
|
|
|
$kmore->( $new_seed, @items ) |
277
|
|
|
|
|
|
|
$kdone->( @items ) |
278
|
|
|
|
|
|
|
|
279
|
|
|
|
|
|
|
$k->( @all_items ) |
280
|
|
|
|
|
|
|
|
281
|
|
|
|
|
|
|
With each iteration, the C is invoked and passed the current C<$seed> |
282
|
|
|
|
|
|
|
value and two continuations, C<$kmore> and C<$kdone>. If C<$kmore> is invoked, |
283
|
|
|
|
|
|
|
the passed items, if any, are appended to the eventual result list. The |
284
|
|
|
|
|
|
|
C is then re-invoked with the new C<$seed> value. If C<$klast> is |
285
|
|
|
|
|
|
|
invoked, the passed items, if any, are appended to the return list, then the |
286
|
|
|
|
|
|
|
entire list is passed to C<$k>. |
287
|
|
|
|
|
|
|
|
288
|
|
|
|
|
|
|
=cut |
289
|
|
|
|
|
|
|
|
290
|
|
|
|
|
|
|
sub gkunfold |
291
|
|
|
|
|
|
|
{ |
292
|
1
|
|
|
1
|
0
|
4
|
my ( $gov, $seed, $body, $k ) = @_; |
293
|
|
|
|
|
|
|
|
294
|
1
|
50
|
|
|
|
8
|
ref $body eq "CODE" or croak 'Expected $body as CODE ref'; |
295
|
|
|
|
|
|
|
|
296
|
1
|
|
|
|
|
2
|
my @ret; |
297
|
|
|
|
|
|
|
|
298
|
|
|
|
|
|
|
gkloop( $gov, |
299
|
|
|
|
|
|
|
sub { |
300
|
5
|
|
|
5
|
|
18
|
my ( $knext, $klast ) = @_; |
301
|
|
|
|
|
|
|
@_ = ( |
302
|
|
|
|
|
|
|
$seed, |
303
|
4
|
|
|
|
|
22
|
sub { $seed = shift; push @ret, @_; goto &$knext }, |
|
4
|
|
|
|
|
7
|
|
|
4
|
|
|
|
|
11
|
|
304
|
1
|
|
|
|
|
5
|
sub { push @ret, @_; goto &$klast }, |
|
1
|
|
|
|
|
2
|
|
305
|
5
|
|
|
|
|
27
|
); |
306
|
5
|
|
|
|
|
17
|
goto &$body; |
307
|
|
|
|
|
|
|
}, |
308
|
1
|
|
|
1
|
|
4
|
sub { $k->( @ret ) }, |
309
|
1
|
|
|
|
|
12
|
); |
310
|
|
|
|
|
|
|
} |
311
|
|
|
|
|
|
|
|
312
|
|
|
|
|
|
|
CPS::_governate "g$_" => $_ for @CPS_PRIMS; |
313
|
|
|
|
|
|
|
|
314
|
|
|
|
|
|
|
=head1 EXAMPLES |
315
|
|
|
|
|
|
|
|
316
|
|
|
|
|
|
|
The following aren't necessarily examples of code which would be found in real |
317
|
|
|
|
|
|
|
programs, but instead, demonstrations of how to use the above functions as |
318
|
|
|
|
|
|
|
ways of controlling program flow. |
319
|
|
|
|
|
|
|
|
320
|
|
|
|
|
|
|
Without dragging in large amount of detail on an asynchronous or event-driven |
321
|
|
|
|
|
|
|
framework, it is difficult to give a useful example of behaviour that CPS |
322
|
|
|
|
|
|
|
allows that couldn't be done just as easily without. Nevertheless, I hope the |
323
|
|
|
|
|
|
|
following examples will be useful to demonstrate use of the above functions, |
324
|
|
|
|
|
|
|
in a way which hints at their use in a real program. |
325
|
|
|
|
|
|
|
|
326
|
|
|
|
|
|
|
=head2 Implementing C using C |
327
|
|
|
|
|
|
|
|
328
|
|
|
|
|
|
|
use CPS::Functional qw( kfoldl ); |
329
|
|
|
|
|
|
|
|
330
|
|
|
|
|
|
|
my @words = qw( My message here ); |
331
|
|
|
|
|
|
|
|
332
|
|
|
|
|
|
|
kfoldl( |
333
|
|
|
|
|
|
|
\@words, |
334
|
|
|
|
|
|
|
sub { |
335
|
|
|
|
|
|
|
my ( $left, $right, $k ) = @_; |
336
|
|
|
|
|
|
|
|
337
|
|
|
|
|
|
|
$k->( "$left $right" ); |
338
|
|
|
|
|
|
|
}, |
339
|
|
|
|
|
|
|
sub { |
340
|
|
|
|
|
|
|
my ( $str ) = @_; |
341
|
|
|
|
|
|
|
|
342
|
|
|
|
|
|
|
print "Joined up words: $str\n"; |
343
|
|
|
|
|
|
|
} |
344
|
|
|
|
|
|
|
); |
345
|
|
|
|
|
|
|
|
346
|
|
|
|
|
|
|
=head2 Implementing C using C |
347
|
|
|
|
|
|
|
|
348
|
|
|
|
|
|
|
The following program illustrates the way that C can split a |
349
|
|
|
|
|
|
|
string, in a reverse way to the way C can join it. |
350
|
|
|
|
|
|
|
|
351
|
|
|
|
|
|
|
use CPS::Functional qw( kunfold ); |
352
|
|
|
|
|
|
|
|
353
|
|
|
|
|
|
|
my $str = "My message here"; |
354
|
|
|
|
|
|
|
|
355
|
|
|
|
|
|
|
kunfold( |
356
|
|
|
|
|
|
|
$str, |
357
|
|
|
|
|
|
|
sub { |
358
|
|
|
|
|
|
|
my ( $s, $kmore, $kdone ) = @_; |
359
|
|
|
|
|
|
|
|
360
|
|
|
|
|
|
|
if( $s =~ s/^(.*?) // ) { |
361
|
|
|
|
|
|
|
return $kmore->( $s, $1 ); |
362
|
|
|
|
|
|
|
} |
363
|
|
|
|
|
|
|
else { |
364
|
|
|
|
|
|
|
return $kdone->( $s ); |
365
|
|
|
|
|
|
|
} |
366
|
|
|
|
|
|
|
}, |
367
|
|
|
|
|
|
|
sub { |
368
|
|
|
|
|
|
|
my @words = @_; |
369
|
|
|
|
|
|
|
print "Words in message:\n"; |
370
|
|
|
|
|
|
|
print "$_\n" for @words; |
371
|
|
|
|
|
|
|
} |
372
|
|
|
|
|
|
|
); |
373
|
|
|
|
|
|
|
|
374
|
|
|
|
|
|
|
=head2 Generating Prime Numbers |
375
|
|
|
|
|
|
|
|
376
|
|
|
|
|
|
|
While the design of C is symmetric to C, the seed value |
377
|
|
|
|
|
|
|
doesn't have to be successively broken apart into pieces. Another valid use |
378
|
|
|
|
|
|
|
for it may be storing intermediate values in computation, such as in this |
379
|
|
|
|
|
|
|
example, storing a list of known primes, to help generate the next one: |
380
|
|
|
|
|
|
|
|
381
|
|
|
|
|
|
|
use CPS::Functional qw( kunfold ); |
382
|
|
|
|
|
|
|
|
383
|
|
|
|
|
|
|
kunfold( |
384
|
|
|
|
|
|
|
[ 2, 3 ], |
385
|
|
|
|
|
|
|
sub { |
386
|
|
|
|
|
|
|
my ( $vals, $kmore, $kdone ) = @_; |
387
|
|
|
|
|
|
|
|
388
|
|
|
|
|
|
|
return $kdone->() if @$vals >= 50; |
389
|
|
|
|
|
|
|
|
390
|
|
|
|
|
|
|
PRIME: for( my $n = $vals->[-1] + 2; ; $n += 2 ) { |
391
|
|
|
|
|
|
|
$n % $_ == 0 and next PRIME for @$vals; |
392
|
|
|
|
|
|
|
|
393
|
|
|
|
|
|
|
push @$vals, $n; |
394
|
|
|
|
|
|
|
return $kmore->( $vals, $n ); |
395
|
|
|
|
|
|
|
} |
396
|
|
|
|
|
|
|
}, |
397
|
|
|
|
|
|
|
sub { |
398
|
|
|
|
|
|
|
my @primes = ( 2, 3, @_ ); |
399
|
|
|
|
|
|
|
print "Primes are @primes\n"; |
400
|
|
|
|
|
|
|
} |
401
|
|
|
|
|
|
|
); |
402
|
|
|
|
|
|
|
|
403
|
|
|
|
|
|
|
=head2 Forward-reading Program Flow |
404
|
|
|
|
|
|
|
|
405
|
|
|
|
|
|
|
One side benefit of the CPS control-flow methods which is unassociated with |
406
|
|
|
|
|
|
|
asynchronous operation, is that the flow of data reads in a more natural |
407
|
|
|
|
|
|
|
left-to-right direction, instead of the right-to-left flow in functional |
408
|
|
|
|
|
|
|
style. Compare |
409
|
|
|
|
|
|
|
|
410
|
|
|
|
|
|
|
sub square { $_ * $_ } |
411
|
|
|
|
|
|
|
sub add { $a + $b } |
412
|
|
|
|
|
|
|
|
413
|
|
|
|
|
|
|
print reduce( \&add, map( square, primes(10) ) ); |
414
|
|
|
|
|
|
|
|
415
|
|
|
|
|
|
|
(because C |
416
|
|
|
|
|
|
|
prototype, it has a different way to pass in the named functions) |
417
|
|
|
|
|
|
|
|
418
|
|
|
|
|
|
|
with |
419
|
|
|
|
|
|
|
|
420
|
|
|
|
|
|
|
my $ksquare = liftk { $_[0] * $_[0] }; |
421
|
|
|
|
|
|
|
my $kadd = liftk { $_[0] + $_[1] }; |
422
|
|
|
|
|
|
|
|
423
|
|
|
|
|
|
|
kprimes 10, sub { |
424
|
|
|
|
|
|
|
kmap \@_, $ksquare, sub { |
425
|
|
|
|
|
|
|
kfoldl \@_, $kadd, sub { |
426
|
|
|
|
|
|
|
print $_[0]; |
427
|
|
|
|
|
|
|
} |
428
|
|
|
|
|
|
|
} |
429
|
|
|
|
|
|
|
}; |
430
|
|
|
|
|
|
|
|
431
|
|
|
|
|
|
|
This translates roughly to a functional vs imperative way to describe the |
432
|
|
|
|
|
|
|
problem: |
433
|
|
|
|
|
|
|
|
434
|
|
|
|
|
|
|
Print the sum of the squares of the first 10 primes. |
435
|
|
|
|
|
|
|
|
436
|
|
|
|
|
|
|
Take the first 10 primes. Square them. Sum them. Print. |
437
|
|
|
|
|
|
|
|
438
|
|
|
|
|
|
|
Admittedly the closure creation somewhat clouds the point in this small |
439
|
|
|
|
|
|
|
example, but in a larger example, the real problem-solving logic would be |
440
|
|
|
|
|
|
|
larger, and stand out more clearly against the background boilerplate. |
441
|
|
|
|
|
|
|
|
442
|
|
|
|
|
|
|
=head1 SEE ALSO |
443
|
|
|
|
|
|
|
|
444
|
|
|
|
|
|
|
=over 4 |
445
|
|
|
|
|
|
|
|
446
|
|
|
|
|
|
|
=item * |
447
|
|
|
|
|
|
|
|
448
|
|
|
|
|
|
|
L - manage flow of control in Continuation-Passing Style |
449
|
|
|
|
|
|
|
|
450
|
|
|
|
|
|
|
=back |
451
|
|
|
|
|
|
|
|
452
|
|
|
|
|
|
|
=head1 AUTHOR |
453
|
|
|
|
|
|
|
|
454
|
|
|
|
|
|
|
Paul Evans |
455
|
|
|
|
|
|
|
|
456
|
|
|
|
|
|
|
=cut |
457
|
|
|
|
|
|
|
|
458
|
|
|
|
|
|
|
0x55AA; |