| line |
stmt |
bran |
cond |
sub |
pod |
time |
code |
|
1
|
|
|
|
|
|
|
package Math::Prime::FastSieve; |
|
2
|
|
|
|
|
|
|
|
|
3
|
8
|
|
|
8
|
|
121672
|
use 5.008001; |
|
|
8
|
|
|
|
|
23
|
|
|
|
8
|
|
|
|
|
271
|
|
|
4
|
8
|
|
|
8
|
|
30
|
use strict; |
|
|
8
|
|
|
|
|
9
|
|
|
|
8
|
|
|
|
|
247
|
|
|
5
|
8
|
|
|
8
|
|
36
|
use warnings; |
|
|
8
|
|
|
|
|
12
|
|
|
|
8
|
|
|
|
|
565
|
|
|
6
|
|
|
|
|
|
|
|
|
7
|
|
|
|
|
|
|
require Exporter; |
|
8
|
|
|
|
|
|
|
|
|
9
|
|
|
|
|
|
|
our @ISA = qw(Exporter); ## no critic (isa) |
|
10
|
|
|
|
|
|
|
|
|
11
|
|
|
|
|
|
|
our @EXPORT_OK = qw( primes ); # We can export primes(). |
|
12
|
|
|
|
|
|
|
our @EXPORT = qw( ); # Export nothing by default. |
|
13
|
|
|
|
|
|
|
|
|
14
|
|
|
|
|
|
|
# our $VERSION = 'x.xx'; |
|
15
|
|
|
|
|
|
|
|
|
16
|
8
|
|
|
8
|
|
2604
|
use Math::Prime::FastSieve::Inline CPP => 'DATA'; |
|
|
8
|
|
|
|
|
16
|
|
|
|
8
|
|
|
|
|
1125
|
|
|
17
|
|
|
|
|
|
|
|
|
18
|
|
|
|
|
|
|
# No real code here. Everything is implemented in pure C++ using |
|
19
|
|
|
|
|
|
|
# Inline::CPP. |
|
20
|
|
|
|
|
|
|
|
|
21
|
|
|
|
|
|
|
# I've grown tired of mistyping isprime() as is_prime(). |
|
22
|
|
|
|
|
|
|
*Math::Prime::FastSieve::Sieve::is_prime |
|
23
|
|
|
|
|
|
|
= \&Math::Prime::FastSieve::Sieve::isprime; |
|
24
|
|
|
|
|
|
|
|
|
25
|
|
|
|
|
|
|
1; |
|
26
|
|
|
|
|
|
|
|
|
27
|
|
|
|
|
|
|
=head1 NAME |
|
28
|
|
|
|
|
|
|
|
|
29
|
|
|
|
|
|
|
Math::Prime::FastSieve - Generate a list of all primes less than or equal |
|
30
|
|
|
|
|
|
|
to C<$n>. Do it quickly. |
|
31
|
|
|
|
|
|
|
|
|
32
|
|
|
|
|
|
|
While we're at it, supply a few additional tools that a Prime Sieve |
|
33
|
|
|
|
|
|
|
facilitates. |
|
34
|
|
|
|
|
|
|
|
|
35
|
|
|
|
|
|
|
This is an "Alt::" distribution that stands in place of the original. |
|
36
|
|
|
|
|
|
|
|
|
37
|
|
|
|
|
|
|
=head1 DESCRIPTION |
|
38
|
|
|
|
|
|
|
|
|
39
|
|
|
|
|
|
|
This module provides an optimized implementation of the Sieve of |
|
40
|
|
|
|
|
|
|
Eratosthenes, and uses it to return a reference to an array all primes up to |
|
41
|
|
|
|
|
|
|
any integer specified, within the limitations of addressable memory. |
|
42
|
|
|
|
|
|
|
|
|
43
|
|
|
|
|
|
|
Additionally the module provides access to other Prime-related functions |
|
44
|
|
|
|
|
|
|
that are facilitated as a by-product of having a really fast Prime |
|
45
|
|
|
|
|
|
|
Sieve. |
|
46
|
|
|
|
|
|
|
|
|
47
|
|
|
|
|
|
|
At the time of writing, the C function will return all primes |
|
48
|
|
|
|
|
|
|
up to and including C<$n> faster than any other module I can find on CPAN |
|
49
|
|
|
|
|
|
|
(including Math::Prime::XS). While a segmented sieve (which this isn't) would |
|
50
|
|
|
|
|
|
|
extend the range of primes accessible, the fact that this module uses |
|
51
|
|
|
|
|
|
|
a bit-sieve means that primes over a billion are easily within reach |
|
52
|
|
|
|
|
|
|
of most modern systems. |
|
53
|
|
|
|
|
|
|
|
|
54
|
|
|
|
|
|
|
|
|
55
|
|
|
|
|
|
|
=head1 SYNOPSIS |
|
56
|
|
|
|
|
|
|
|
|
57
|
|
|
|
|
|
|
# ---------- The simple interface: ---------- |
|
58
|
|
|
|
|
|
|
use Math::Prime::FastSieve qw( primes ); |
|
59
|
|
|
|
|
|
|
|
|
60
|
|
|
|
|
|
|
# Obtain an reference to an array containing all primes less than or |
|
61
|
|
|
|
|
|
|
# equal to 5 Million. |
|
62
|
|
|
|
|
|
|
my $aref = primes( 5_000_000 ); |
|
63
|
|
|
|
|
|
|
|
|
64
|
|
|
|
|
|
|
|
|
65
|
|
|
|
|
|
|
|
|
66
|
|
|
|
|
|
|
# ---------- The object (far more flexible) interface. ---------- |
|
67
|
|
|
|
|
|
|
|
|
68
|
|
|
|
|
|
|
# Create a new sieve and flag all primes less or equal to n. |
|
69
|
|
|
|
|
|
|
my $sieve = Math::Prime::FastSieve::Sieve->new( 5_000_000 ); |
|
70
|
|
|
|
|
|
|
|
|
71
|
|
|
|
|
|
|
# Obtain a ref to an array containing all primes <= 5 Million. |
|
72
|
|
|
|
|
|
|
my $aref = $sieve->primes( 5_000_000 ); |
|
73
|
|
|
|
|
|
|
|
|
74
|
|
|
|
|
|
|
# Obtain a ref to an array containing all primes >= 5, and <= 16. |
|
75
|
|
|
|
|
|
|
my $aref = $sieve->ranged_primes( 5, 16 ); |
|
76
|
|
|
|
|
|
|
|
|
77
|
|
|
|
|
|
|
# Query the sieve: Is 17 prime? Return a true or false value. |
|
78
|
|
|
|
|
|
|
my $result = $sieve->isprime( 17 ); |
|
79
|
|
|
|
|
|
|
|
|
80
|
|
|
|
|
|
|
# Get the value of the nearest prime less than or equal to 42. |
|
81
|
|
|
|
|
|
|
my $less_or_equal = $sieve->nearest_le( 42 ); |
|
82
|
|
|
|
|
|
|
|
|
83
|
|
|
|
|
|
|
# Get the value of the nearest prime greater than or equal to 42. |
|
84
|
|
|
|
|
|
|
my $greater_or_equal = $sieve->nearest_ge( 42 ); |
|
85
|
|
|
|
|
|
|
|
|
86
|
|
|
|
|
|
|
# Count how many primes exist within the sieve (ie, count all primes less |
|
87
|
|
|
|
|
|
|
# than or equal to 5 Million, assuming we instantiated the sieve with |
|
88
|
|
|
|
|
|
|
# Math::Prime::FastSieve::Sieve->new( 5_000_000 );. |
|
89
|
|
|
|
|
|
|
my $num_found = $sieve->count_sieve(); |
|
90
|
|
|
|
|
|
|
|
|
91
|
|
|
|
|
|
|
# Count how many primes fall between 1 and 42 inclusive. |
|
92
|
|
|
|
|
|
|
my $quantity_le = $sieve->count_le( 42 ); |
|
93
|
|
|
|
|
|
|
|
|
94
|
|
|
|
|
|
|
# Return the value of the 42nd prime number. |
|
95
|
|
|
|
|
|
|
my $fourty_second_prime = $sieve->nth_prime( 42 ); |
|
96
|
|
|
|
|
|
|
|
|
97
|
|
|
|
|
|
|
|
|
98
|
|
|
|
|
|
|
=head1 EXPORT |
|
99
|
|
|
|
|
|
|
|
|
100
|
|
|
|
|
|
|
Exports nothing by default. If you ask nicely it will export the single |
|
101
|
|
|
|
|
|
|
subroutine, C. |
|
102
|
|
|
|
|
|
|
|
|
103
|
|
|
|
|
|
|
use Math::Prime::FastSieve qw( primes ); # Import primes(). |
|
104
|
|
|
|
|
|
|
use Math::Prime::FastSieve; # No imports. |
|
105
|
|
|
|
|
|
|
|
|
106
|
|
|
|
|
|
|
=head1 SUBROUTINES/METHODS |
|
107
|
|
|
|
|
|
|
|
|
108
|
|
|
|
|
|
|
This module provides two modus operandi. The simplest also happens to |
|
109
|
|
|
|
|
|
|
be the fastest way of generating a list of all primes up to and |
|
110
|
|
|
|
|
|
|
including C<$n>. That is via a direct call to the C |
|
111
|
|
|
|
|
|
|
function. |
|
112
|
|
|
|
|
|
|
|
|
113
|
|
|
|
|
|
|
The more powerful way to use this module is via its object oriented |
|
114
|
|
|
|
|
|
|
interface. With that approach, the constructor C creates a |
|
115
|
|
|
|
|
|
|
prime sieve flagging all primes from 2 to C<$n> inclusive, and returns |
|
116
|
|
|
|
|
|
|
a Sieve object. That object may then be queried by way of accessor |
|
117
|
|
|
|
|
|
|
methods to get at any of the following: |
|
118
|
|
|
|
|
|
|
|
|
119
|
|
|
|
|
|
|
=over 4 |
|
120
|
|
|
|
|
|
|
|
|
121
|
|
|
|
|
|
|
=item * C |
|
122
|
|
|
|
|
|
|
A list of all primes within the sieve. |
|
123
|
|
|
|
|
|
|
|
|
124
|
|
|
|
|
|
|
=item * C |
|
125
|
|
|
|
|
|
|
A list of all primes >= C<$lower>, and <= C<$upper>. |
|
126
|
|
|
|
|
|
|
|
|
127
|
|
|
|
|
|
|
=item * C |
|
128
|
|
|
|
|
|
|
=item * C |
|
129
|
|
|
|
|
|
|
A primality test for any integer within the bounds of the sieve. C |
|
130
|
|
|
|
|
|
|
is synonymous for C, mostly because I got tired of forgetting |
|
131
|
|
|
|
|
|
|
how to spell the method. |
|
132
|
|
|
|
|
|
|
|
|
133
|
|
|
|
|
|
|
=item * C |
|
134
|
|
|
|
|
|
|
Find the nearest prime less than or equal to C<$n> within the bounds of |
|
135
|
|
|
|
|
|
|
the sieve. |
|
136
|
|
|
|
|
|
|
|
|
137
|
|
|
|
|
|
|
=item * C |
|
138
|
|
|
|
|
|
|
Find the nearest prime greater or equal to C<$n> within the bounds of |
|
139
|
|
|
|
|
|
|
the sieve. |
|
140
|
|
|
|
|
|
|
|
|
141
|
|
|
|
|
|
|
=item * C |
|
142
|
|
|
|
|
|
|
A count of all primes in the sieve. |
|
143
|
|
|
|
|
|
|
|
|
144
|
|
|
|
|
|
|
=item * C |
|
145
|
|
|
|
|
|
|
A a count of all primes less than or equal to C<$n> within the bounds of |
|
146
|
|
|
|
|
|
|
the sieve. |
|
147
|
|
|
|
|
|
|
|
|
148
|
|
|
|
|
|
|
=item * C |
|
149
|
|
|
|
|
|
|
Return the C<$n>th prime, within the bounds of the sieve. |
|
150
|
|
|
|
|
|
|
|
|
151
|
|
|
|
|
|
|
=back |
|
152
|
|
|
|
|
|
|
|
|
153
|
|
|
|
|
|
|
Because the sieve is created when the object is instantiated, many of |
|
154
|
|
|
|
|
|
|
the tests you might call on the sieve object will execute quite quickly. |
|
155
|
|
|
|
|
|
|
Each of the subs and methods documented below will also attempt to describe |
|
156
|
|
|
|
|
|
|
the computational and memory complexity of the function. |
|
157
|
|
|
|
|
|
|
|
|
158
|
|
|
|
|
|
|
=head1 The Standard Interface |
|
159
|
|
|
|
|
|
|
|
|
160
|
|
|
|
|
|
|
=head2 Objective |
|
161
|
|
|
|
|
|
|
|
|
162
|
|
|
|
|
|
|
Provide a fast and simple means of generating a big list of prime |
|
163
|
|
|
|
|
|
|
numbers. |
|
164
|
|
|
|
|
|
|
|
|
165
|
|
|
|
|
|
|
=head3 primes() |
|
166
|
|
|
|
|
|
|
|
|
167
|
|
|
|
|
|
|
This is a regular function (ie, not an object method). |
|
168
|
|
|
|
|
|
|
|
|
169
|
|
|
|
|
|
|
Pass a positive integer as its parameter. Returns a reference to an |
|
170
|
|
|
|
|
|
|
array of all prime numbers C<2 .. $n>. |
|
171
|
|
|
|
|
|
|
|
|
172
|
|
|
|
|
|
|
This function is implemented in C++ using Inline::CPP, which in turn |
|
173
|
|
|
|
|
|
|
binds it to Perl via XS. The method used is the Sieve of Eratosthenes. |
|
174
|
|
|
|
|
|
|
The sieve is one of the fastest known methods of generating a list of |
|
175
|
|
|
|
|
|
|
primes up to C<$n>. |
|
176
|
|
|
|
|
|
|
|
|
177
|
|
|
|
|
|
|
This implementation makes several optimizations beyond the cannonical |
|
178
|
|
|
|
|
|
|
Sieve, including: |
|
179
|
|
|
|
|
|
|
|
|
180
|
|
|
|
|
|
|
=over 4 |
|
181
|
|
|
|
|
|
|
|
|
182
|
|
|
|
|
|
|
=item * Uses a bit vector as a sieve: A memory optimization that allows |
|
183
|
|
|
|
|
|
|
the sieve to scale well without consuming so much memory that your |
|
184
|
|
|
|
|
|
|
system grinds to a stand-still for large C<$n>s (where "large" depends |
|
185
|
|
|
|
|
|
|
on your system, of course). |
|
186
|
|
|
|
|
|
|
|
|
187
|
|
|
|
|
|
|
=item * The sieve's bits only represent odd numbers (memory optimization). |
|
188
|
|
|
|
|
|
|
|
|
189
|
|
|
|
|
|
|
=item * Only sifts and tests odd integers. (2 is handled as a special case.) |
|
190
|
|
|
|
|
|
|
|
|
191
|
|
|
|
|
|
|
=item * Returns an array-ref rather than a potentially huge (slow) list. |
|
192
|
|
|
|
|
|
|
|
|
193
|
|
|
|
|
|
|
=item * Other micro-optimizations to squeeze a few cycles out here |
|
194
|
|
|
|
|
|
|
and there. |
|
195
|
|
|
|
|
|
|
|
|
196
|
|
|
|
|
|
|
=back |
|
197
|
|
|
|
|
|
|
|
|
198
|
|
|
|
|
|
|
The result is a prime number generator that is...fast. It operates in |
|
199
|
|
|
|
|
|
|
O( n log log n ) time, with a O(n) memory growth rate. |
|
200
|
|
|
|
|
|
|
|
|
201
|
|
|
|
|
|
|
=head1 The Object Oriented Interface |
|
202
|
|
|
|
|
|
|
|
|
203
|
|
|
|
|
|
|
=head2 Objective |
|
204
|
|
|
|
|
|
|
|
|
205
|
|
|
|
|
|
|
Where the standard interface wraps the sieve constructor and the sieve |
|
206
|
|
|
|
|
|
|
accessor in a single function called C, the object oriented |
|
207
|
|
|
|
|
|
|
interface places the sieve constructor in |
|
208
|
|
|
|
|
|
|
Cnew()>, and leaves the interpretation |
|
209
|
|
|
|
|
|
|
of the sieve's results to separate accessor methods. The advantage is |
|
210
|
|
|
|
|
|
|
that if you don't really need "a big list", but instead need other |
|
211
|
|
|
|
|
|
|
functionality that may be computationally (and memory) cheaper to obtain |
|
212
|
|
|
|
|
|
|
from the sieve, you can get at those results without constructing a huge |
|
213
|
|
|
|
|
|
|
list of primes. |
|
214
|
|
|
|
|
|
|
|
|
215
|
|
|
|
|
|
|
The standard interface is slightly faster if you just want a big list. But |
|
216
|
|
|
|
|
|
|
the object interface is still very fast, and provides greater flexibility, |
|
217
|
|
|
|
|
|
|
including the ability to use C to process primes in smaller, |
|
218
|
|
|
|
|
|
|
more memory efficient chunks. |
|
219
|
|
|
|
|
|
|
|
|
220
|
|
|
|
|
|
|
|
|
221
|
|
|
|
|
|
|
=head3 new() |
|
222
|
|
|
|
|
|
|
|
|
223
|
|
|
|
|
|
|
Class method of C Requires a single |
|
224
|
|
|
|
|
|
|
integer parameter that represents the largest value to be held in the |
|
225
|
|
|
|
|
|
|
sieve. For example: |
|
226
|
|
|
|
|
|
|
|
|
227
|
|
|
|
|
|
|
my $sieve = Math::Prime::FastSieve::Sieve->new( 1_000_000_000 ); |
|
228
|
|
|
|
|
|
|
|
|
229
|
|
|
|
|
|
|
This will create a Sieve object that flags all integers from 2 to |
|
230
|
|
|
|
|
|
|
1 billion that are prime. |
|
231
|
|
|
|
|
|
|
|
|
232
|
|
|
|
|
|
|
Calling C is an O(n log log n) operation. The memory growth is at a |
|
233
|
|
|
|
|
|
|
rate that is 1/8th the rate of growth of C<$n>. |
|
234
|
|
|
|
|
|
|
|
|
235
|
|
|
|
|
|
|
|
|
236
|
|
|
|
|
|
|
=head3 primes() |
|
237
|
|
|
|
|
|
|
|
|
238
|
|
|
|
|
|
|
This works just like the standard C function, except that it |
|
239
|
|
|
|
|
|
|
is a member function of your Sieve object, and also (behind the scenes) |
|
240
|
|
|
|
|
|
|
it doesn't need to create a sieve of prime flags since C already |
|
241
|
|
|
|
|
|
|
did that for you. You must call C with an integer parameter. |
|
242
|
|
|
|
|
|
|
The integer must be less than or equal to the integer value previously |
|
243
|
|
|
|
|
|
|
passed to the C constructor. C returns a reference to |
|
244
|
|
|
|
|
|
|
an array of all prime numbers from 2 to the integer passed to it. |
|
245
|
|
|
|
|
|
|
|
|
246
|
|
|
|
|
|
|
Passing an out-of-bounds integer will result in a return value of an |
|
247
|
|
|
|
|
|
|
array ref pointing to an empty array. |
|
248
|
|
|
|
|
|
|
|
|
249
|
|
|
|
|
|
|
my $primes_ref = $sieve->primes( 1_000_000_000 ); |
|
250
|
|
|
|
|
|
|
my $primes_ref = $sieve->primes( 50 ); |
|
251
|
|
|
|
|
|
|
|
|
252
|
|
|
|
|
|
|
The C method is an O(n) operation for both time and memory. |
|
253
|
|
|
|
|
|
|
|
|
254
|
|
|
|
|
|
|
=head3 ranged_primes() |
|
255
|
|
|
|
|
|
|
|
|
256
|
|
|
|
|
|
|
This behaves exactly like the C method, except that you |
|
257
|
|
|
|
|
|
|
specify a lower and upper limit within the bounds of the sieve. The |
|
258
|
|
|
|
|
|
|
return value is a reference to an array holding all prime numbers |
|
259
|
|
|
|
|
|
|
greater than or equal to the lower limit, and less than or equal to the |
|
260
|
|
|
|
|
|
|
upper limit. |
|
261
|
|
|
|
|
|
|
|
|
262
|
|
|
|
|
|
|
The purpose of this method is to allow you to create a sieve ( with |
|
263
|
|
|
|
|
|
|
C ), and then get results in a segmented form. The reasons this |
|
264
|
|
|
|
|
|
|
may be desirable are two-fold: First, you may only need a subset. |
|
265
|
|
|
|
|
|
|
Second, this gives you finer control over how much memory is gobbled up |
|
266
|
|
|
|
|
|
|
by the list returned. For a huge sieve the sieve's memory footprint is |
|
267
|
|
|
|
|
|
|
much smaller than the list of integers that are flagged as prime. By |
|
268
|
|
|
|
|
|
|
dealing with that list in chunks you can have a sieve of billions of |
|
269
|
|
|
|
|
|
|
prime flags, but never hold that big of a list of integers in memory |
|
270
|
|
|
|
|
|
|
all at once. |
|
271
|
|
|
|
|
|
|
|
|
272
|
|
|
|
|
|
|
my $primes_ref = $sieve->ranged_primes( 5, 16 ); |
|
273
|
|
|
|
|
|
|
# $primes_ref now holds [ 5, 7, 11, 13 ]. |
|
274
|
|
|
|
|
|
|
|
|
275
|
|
|
|
|
|
|
The time complexity of this method is O(n) where 'n' is the upper limit |
|
276
|
|
|
|
|
|
|
minus the lower limit. So a range of 5_000_000 .. 5_000_010 consumes as |
|
277
|
|
|
|
|
|
|
much time as 100 .. 110. |
|
278
|
|
|
|
|
|
|
|
|
279
|
|
|
|
|
|
|
|
|
280
|
|
|
|
|
|
|
=head3 isprime() |
|
281
|
|
|
|
|
|
|
=head3 is_prime() |
|
282
|
|
|
|
|
|
|
|
|
283
|
|
|
|
|
|
|
Pass a parameter consisiting of a single integer within the range of the |
|
284
|
|
|
|
|
|
|
Sieve object. Returns true if the integer is prime, false otherwise. |
|
285
|
|
|
|
|
|
|
If the integer is out of the Sieve object's bounds, the result will be |
|
286
|
|
|
|
|
|
|
false. |
|
287
|
|
|
|
|
|
|
|
|
288
|
|
|
|
|
|
|
if( $sieve->isprime(42) ) { |
|
289
|
|
|
|
|
|
|
say "42 is prime."; |
|
290
|
|
|
|
|
|
|
} else { |
|
291
|
|
|
|
|
|
|
say "42 isn't prime."; |
|
292
|
|
|
|
|
|
|
} |
|
293
|
|
|
|
|
|
|
|
|
294
|
|
|
|
|
|
|
C is a synonym for C. They're the same method; I |
|
295
|
|
|
|
|
|
|
just grew tired of forgetting which spelling to use when calling the |
|
296
|
|
|
|
|
|
|
method. |
|
297
|
|
|
|
|
|
|
|
|
298
|
|
|
|
|
|
|
This is an O(1) operation. |
|
299
|
|
|
|
|
|
|
|
|
300
|
|
|
|
|
|
|
|
|
301
|
|
|
|
|
|
|
=head3 nearest_le() |
|
302
|
|
|
|
|
|
|
|
|
303
|
|
|
|
|
|
|
The C method returns the closest prime that is less than |
|
304
|
|
|
|
|
|
|
or equal to its integer parameter. Passing an out of bounds parameter |
|
305
|
|
|
|
|
|
|
will return a C<0>. |
|
306
|
|
|
|
|
|
|
|
|
307
|
|
|
|
|
|
|
my $nearest_less_or_equal = $sieve->nearest_le( 42 ); |
|
308
|
|
|
|
|
|
|
|
|
309
|
|
|
|
|
|
|
Since the nearest prime is never very far away, this is an |
|
310
|
|
|
|
|
|
|
O( n / ( log n ) ) operation. |
|
311
|
|
|
|
|
|
|
|
|
312
|
|
|
|
|
|
|
|
|
313
|
|
|
|
|
|
|
=head3 nearest_ge() |
|
314
|
|
|
|
|
|
|
|
|
315
|
|
|
|
|
|
|
Like the C method, but this method returns the prime that |
|
316
|
|
|
|
|
|
|
is greater than or equal to the input parameter. If the input param. is |
|
317
|
|
|
|
|
|
|
out of range, or if the next prime is out of range, a C<0> is returned. |
|
318
|
|
|
|
|
|
|
|
|
319
|
|
|
|
|
|
|
my $nearest_greater_or_equal = $sieve->nearest_ge( 42 ); |
|
320
|
|
|
|
|
|
|
|
|
321
|
|
|
|
|
|
|
This is also an O( n / ( log n ) ) operation. |
|
322
|
|
|
|
|
|
|
|
|
323
|
|
|
|
|
|
|
By adding one to the return value and passing that new value as a |
|
324
|
|
|
|
|
|
|
parameter to the C method again and again in a loop it is |
|
325
|
|
|
|
|
|
|
easy to iterate through a list of primes without generating them all |
|
326
|
|
|
|
|
|
|
at once. Of course it's not going to be as fast as getting the big |
|
327
|
|
|
|
|
|
|
list all at once, but you can't have everything in life, now can you? |
|
328
|
|
|
|
|
|
|
|
|
329
|
|
|
|
|
|
|
|
|
330
|
|
|
|
|
|
|
=head3 count_sieve() |
|
331
|
|
|
|
|
|
|
|
|
332
|
|
|
|
|
|
|
Takes no input parameter. Counts all of the primes in the sieve and |
|
333
|
|
|
|
|
|
|
returns the count. The first time this is called on a Sieve object |
|
334
|
|
|
|
|
|
|
the count takes O(n) time. Subsequent calls benefit from the first |
|
335
|
|
|
|
|
|
|
run being cached, and thus become O(1) time. |
|
336
|
|
|
|
|
|
|
|
|
337
|
|
|
|
|
|
|
|
|
338
|
|
|
|
|
|
|
=head3 count_le() |
|
339
|
|
|
|
|
|
|
|
|
340
|
|
|
|
|
|
|
Pass an integer within the range of the sieve as a parameter. Return |
|
341
|
|
|
|
|
|
|
value is a count of how many primes are less than or equal to that |
|
342
|
|
|
|
|
|
|
integer. If the value passed as a parameter is the same as the size of |
|
343
|
|
|
|
|
|
|
the sieve, the results are cached for future calls to C. |
|
344
|
|
|
|
|
|
|
|
|
345
|
|
|
|
|
|
|
This is an O(n) operation. |
|
346
|
|
|
|
|
|
|
|
|
347
|
|
|
|
|
|
|
|
|
348
|
|
|
|
|
|
|
=head3 nth_prime() |
|
349
|
|
|
|
|
|
|
|
|
350
|
|
|
|
|
|
|
This method returns the n-th prime, where C<$n> is the cardinal index in the |
|
351
|
|
|
|
|
|
|
sequence of primes. For example: |
|
352
|
|
|
|
|
|
|
|
|
353
|
|
|
|
|
|
|
say $sieve->nth_prime(1) # prints 2: the first prime is 2. |
|
354
|
|
|
|
|
|
|
say $sieve->nth_prime(3) # prints 5: the third prime is 5. |
|
355
|
|
|
|
|
|
|
|
|
356
|
|
|
|
|
|
|
If there is no nth prime in the bounds of the sieve C<0> is returned. |
|
357
|
|
|
|
|
|
|
|
|
358
|
|
|
|
|
|
|
|
|
359
|
|
|
|
|
|
|
|
|
360
|
|
|
|
|
|
|
|
|
361
|
|
|
|
|
|
|
=head1 Implementation Notes |
|
362
|
|
|
|
|
|
|
|
|
363
|
|
|
|
|
|
|
The sieve is implemented as a bit sieve using a C++ vector. All odd |
|
364
|
|
|
|
|
|
|
integers from 3 to C<$n> are represented based on their index within the |
|
365
|
|
|
|
|
|
|
sieve. The only even prime, 2, is handled as a special case. A bit sieve is |
|
366
|
|
|
|
|
|
|
highly efficient from a memory standpoint because obviously it only consumes |
|
367
|
|
|
|
|
|
|
one byte per eight integers. This sieve is further optimized by reperesenting |
|
368
|
|
|
|
|
|
|
only odd integers in the sieve. So a sieve from 0 .. 1_000_000_000 only needs |
|
369
|
|
|
|
|
|
|
500_000_000 bits, or 59.6 MB. |
|
370
|
|
|
|
|
|
|
|
|
371
|
|
|
|
|
|
|
While a bit sieve was used for memory efficiency, just about every |
|
372
|
|
|
|
|
|
|
other optimization favored reducing time complexity. |
|
373
|
|
|
|
|
|
|
|
|
374
|
|
|
|
|
|
|
Furthermore, all functions and methods are implemented in C++ by way of |
|
375
|
|
|
|
|
|
|
Inline::CPP. In fact, the sieve itself is never exposed to Perl (this |
|
376
|
|
|
|
|
|
|
decision is both a time and memory optimization). |
|
377
|
|
|
|
|
|
|
|
|
378
|
|
|
|
|
|
|
A side effect of using a bit sieve is that the sieve itself actually |
|
379
|
|
|
|
|
|
|
requires far less memory than the integer list of primes sifted from it. |
|
380
|
|
|
|
|
|
|
That means that the memory bottleneck with the C function, as |
|
381
|
|
|
|
|
|
|
well as with the C object method is not, in fact, the sieve, |
|
382
|
|
|
|
|
|
|
but the list passed back to Perl via an array-ref. |
|
383
|
|
|
|
|
|
|
|
|
384
|
|
|
|
|
|
|
If you find that your system memory isn't allowing you to call C |
|
385
|
|
|
|
|
|
|
with as big an integer as you wish, use the object oriented interface. |
|
386
|
|
|
|
|
|
|
C will generate a sieve up to the largest integer your system. Then |
|
387
|
|
|
|
|
|
|
rather than calling the C method, use C, or |
|
388
|
|
|
|
|
|
|
C to iterate over the list. Of course this is slightly slower, but |
|
389
|
|
|
|
|
|
|
it beats running out of memory doesn't it? |
|
390
|
|
|
|
|
|
|
|
|
391
|
|
|
|
|
|
|
NOTE: As of version 0.10, support for long ints is built into |
|
392
|
|
|
|
|
|
|
Math::Prime::FastSieve. If your version of Perl was built with support for |
|
393
|
|
|
|
|
|
|
long ints, you can create sieve sizes well over the 2.14 billion limit that |
|
394
|
|
|
|
|
|
|
standard ints impose. |
|
395
|
|
|
|
|
|
|
|
|
396
|
|
|
|
|
|
|
=head1 DEPENDENCIES |
|
397
|
|
|
|
|
|
|
|
|
398
|
|
|
|
|
|
|
You will need: L, L (which is packaged |
|
399
|
|
|
|
|
|
|
with Inline), L, |
|
400
|
|
|
|
|
|
|
L, |
|
401
|
|
|
|
|
|
|
L (core), and |
|
402
|
|
|
|
|
|
|
L (core). |
|
403
|
|
|
|
|
|
|
|
|
404
|
|
|
|
|
|
|
=head1 CONFIGURATION AND ENVIRONMENT |
|
405
|
|
|
|
|
|
|
|
|
406
|
|
|
|
|
|
|
In addition to the module dependencies listed in the previous section, your |
|
407
|
|
|
|
|
|
|
system must have a C++ compiler that is compatible with the compiler used to |
|
408
|
|
|
|
|
|
|
build Perl. You may need to install one. Additionally, if your Perl was |
|
409
|
|
|
|
|
|
|
built with long integer support, this module will take advantage of that |
|
410
|
|
|
|
|
|
|
support. |
|
411
|
|
|
|
|
|
|
|
|
412
|
|
|
|
|
|
|
While it may sound like there are a lot of dependencies and configuration |
|
413
|
|
|
|
|
|
|
requirements, in practice, most users should be able to install this module |
|
414
|
|
|
|
|
|
|
with the familiar incantations: |
|
415
|
|
|
|
|
|
|
|
|
416
|
|
|
|
|
|
|
cpan Math::Prime::FastSieve |
|
417
|
|
|
|
|
|
|
|
|
418
|
|
|
|
|
|
|
Or download and unpack the tarball, and... |
|
419
|
|
|
|
|
|
|
|
|
420
|
|
|
|
|
|
|
perl Makefile.PL |
|
421
|
|
|
|
|
|
|
make |
|
422
|
|
|
|
|
|
|
make test |
|
423
|
|
|
|
|
|
|
make install |
|
424
|
|
|
|
|
|
|
|
|
425
|
|
|
|
|
|
|
Using the cpan shell, cpanm, or cpanplus will have the added benefit of |
|
426
|
|
|
|
|
|
|
pulling in and building the dependencies automatically. |
|
427
|
|
|
|
|
|
|
|
|
428
|
|
|
|
|
|
|
=head1 DIAGNOSTICS |
|
429
|
|
|
|
|
|
|
|
|
430
|
|
|
|
|
|
|
If you encounter an installation failure, please email me a transcript of the |
|
431
|
|
|
|
|
|
|
session. |
|
432
|
|
|
|
|
|
|
|
|
433
|
|
|
|
|
|
|
=head1 INCOMPATIBILITIES |
|
434
|
|
|
|
|
|
|
|
|
435
|
|
|
|
|
|
|
This module can only be built using systems that have a C++ compiler, and that |
|
436
|
|
|
|
|
|
|
are able to cleanly install L. That is going to |
|
437
|
|
|
|
|
|
|
cover the majority of potential users. If you're one of the unlucky few, |
|
438
|
|
|
|
|
|
|
please send me an email. Since I also maintain Inline::CPP, your feeback |
|
439
|
|
|
|
|
|
|
may help me to sort out the few remaining installation problems with that |
|
440
|
|
|
|
|
|
|
great module. |
|
441
|
|
|
|
|
|
|
|
|
442
|
|
|
|
|
|
|
=head1 AUTHOR |
|
443
|
|
|
|
|
|
|
|
|
444
|
|
|
|
|
|
|
David Oswald, C<< >> |
|
445
|
|
|
|
|
|
|
|
|
446
|
|
|
|
|
|
|
=head1 BUGS AND LIMITATIONS |
|
447
|
|
|
|
|
|
|
|
|
448
|
|
|
|
|
|
|
Since the Sieve of Eratosthenes requires one 'bit' per integer in the sieve, |
|
449
|
|
|
|
|
|
|
the memory requirements can be high for large tests. Additionally, as the |
|
450
|
|
|
|
|
|
|
result set is generated, because Perl's scalars take up a lot more space than |
|
451
|
|
|
|
|
|
|
one bit, it's even more likely the entire result set won't fit into memory. |
|
452
|
|
|
|
|
|
|
|
|
453
|
|
|
|
|
|
|
The OO interface does provide functions that don't build a big huge |
|
454
|
|
|
|
|
|
|
memory-hungry list. |
|
455
|
|
|
|
|
|
|
|
|
456
|
|
|
|
|
|
|
|
|
457
|
|
|
|
|
|
|
Please report any bugs or feature requests to |
|
458
|
|
|
|
|
|
|
C, or through the web interface |
|
459
|
|
|
|
|
|
|
at L. |
|
460
|
|
|
|
|
|
|
I will be notified, and then you'll automatically be notified of |
|
461
|
|
|
|
|
|
|
progress on your bug as I make changes. |
|
462
|
|
|
|
|
|
|
|
|
463
|
|
|
|
|
|
|
|
|
464
|
|
|
|
|
|
|
|
|
465
|
|
|
|
|
|
|
|
|
466
|
|
|
|
|
|
|
=head1 SUPPORT |
|
467
|
|
|
|
|
|
|
|
|
468
|
|
|
|
|
|
|
You can find documentation for this module with the perldoc command. |
|
469
|
|
|
|
|
|
|
|
|
470
|
|
|
|
|
|
|
perldoc Math::Prime::FastSieve |
|
471
|
|
|
|
|
|
|
|
|
472
|
|
|
|
|
|
|
|
|
473
|
|
|
|
|
|
|
You can also look for information at: |
|
474
|
|
|
|
|
|
|
|
|
475
|
|
|
|
|
|
|
=over 4 |
|
476
|
|
|
|
|
|
|
|
|
477
|
|
|
|
|
|
|
=item * RT: CPAN's request tracker (report bugs here) |
|
478
|
|
|
|
|
|
|
|
|
479
|
|
|
|
|
|
|
L |
|
480
|
|
|
|
|
|
|
|
|
481
|
|
|
|
|
|
|
=item * AnnoCPAN: Annotated CPAN documentation |
|
482
|
|
|
|
|
|
|
|
|
483
|
|
|
|
|
|
|
L |
|
484
|
|
|
|
|
|
|
|
|
485
|
|
|
|
|
|
|
=item * CPAN Ratings |
|
486
|
|
|
|
|
|
|
|
|
487
|
|
|
|
|
|
|
L |
|
488
|
|
|
|
|
|
|
|
|
489
|
|
|
|
|
|
|
=item * Search CPAN |
|
490
|
|
|
|
|
|
|
|
|
491
|
|
|
|
|
|
|
L |
|
492
|
|
|
|
|
|
|
|
|
493
|
|
|
|
|
|
|
=back |
|
494
|
|
|
|
|
|
|
|
|
495
|
|
|
|
|
|
|
|
|
496
|
|
|
|
|
|
|
=head1 ACKNOWLEDGEMENTS |
|
497
|
|
|
|
|
|
|
|
|
498
|
|
|
|
|
|
|
This module is made possible by Inline::CPP, which wouldn't be possible |
|
499
|
|
|
|
|
|
|
without the hard work of the folks involved in the Inline and Inline::C |
|
500
|
|
|
|
|
|
|
project. There are many individuals who have contributed and continue to |
|
501
|
|
|
|
|
|
|
contribute to the Inline project. I won't name them all here, but they do |
|
502
|
|
|
|
|
|
|
deserve thanks and credit. |
|
503
|
|
|
|
|
|
|
|
|
504
|
|
|
|
|
|
|
Dana Jacobsen provided several optimizations that improved even further on |
|
505
|
|
|
|
|
|
|
the speed and memory performance of this module. Dana's contributions include |
|
506
|
|
|
|
|
|
|
reducing the memory footprint of the bit sieve in half, and trimming cycles by |
|
507
|
|
|
|
|
|
|
cutting in half the number of iterations of an inner loop in the sieve |
|
508
|
|
|
|
|
|
|
generator. This module started fast and got even faster (and more memory |
|
509
|
|
|
|
|
|
|
efficient) with Dana's contributions. |
|
510
|
|
|
|
|
|
|
|
|
511
|
|
|
|
|
|
|
=head1 SEE ALSO |
|
512
|
|
|
|
|
|
|
|
|
513
|
|
|
|
|
|
|
L is a much larger Primes library, by Dana Jacobsen. In |
|
514
|
|
|
|
|
|
|
some cases Math::Prime::FastSieve's object-oriented interface may be a better |
|
515
|
|
|
|
|
|
|
fit, but generally speaking, Dana's module is a little faster, and supercedes |
|
516
|
|
|
|
|
|
|
this module in its functionality and capabilities. |
|
517
|
|
|
|
|
|
|
|
|
518
|
|
|
|
|
|
|
=head1 LICENSE AND COPYRIGHT |
|
519
|
|
|
|
|
|
|
|
|
520
|
|
|
|
|
|
|
Copyright 2011-2014 David Oswald. |
|
521
|
|
|
|
|
|
|
|
|
522
|
|
|
|
|
|
|
This program is free software; you can redistribute it and/or modify it |
|
523
|
|
|
|
|
|
|
under the terms of either: the GNU General Public License as published |
|
524
|
|
|
|
|
|
|
by the Free Software Foundation; or the Artistic License. |
|
525
|
|
|
|
|
|
|
|
|
526
|
|
|
|
|
|
|
See http://dev.perl.org/licenses/ for more information. |
|
527
|
|
|
|
|
|
|
|
|
528
|
|
|
|
|
|
|
|
|
529
|
|
|
|
|
|
|
=cut |
|
530
|
|
|
|
|
|
|
|
|
531
|
|
|
|
|
|
|
|
|
532
|
|
|
|
|
|
|
__DATA__ |