dumb_numb_set 0.0.1
Sign up to get free protection for your applications and to get access to all the features.
- data/lib/dumb_numb_set.rb +127 -0
- metadata +54 -0
@@ -0,0 +1,127 @@
|
|
1
|
+
# ### DumbNumbSet
|
2
|
+
#
|
3
|
+
# A DumbNumbSet is a data structure for a very specific purpose: compactly
|
4
|
+
# storing a set of mostly consecutive positive integers that may begin at
|
5
|
+
# any number.
|
6
|
+
#
|
7
|
+
# In Ruby, a Set is actually stored as a Hash with `true` as values. So the
|
8
|
+
# fairest comparison for DumbNumbSet is the same Hash a Set would use.
|
9
|
+
#
|
10
|
+
# #### Usage
|
11
|
+
#
|
12
|
+
# The API for DumbNumbSet is very simple. Numbers can be added, removed, and
|
13
|
+
# their presence in the set can be queried.
|
14
|
+
#
|
15
|
+
# dns = DumbNumbSet.new
|
16
|
+
# dns.add 1
|
17
|
+
# dns.include? 1 #=> true
|
18
|
+
# dns.remove 1
|
19
|
+
# dns.include? 1 #=> false
|
20
|
+
#
|
21
|
+
# #### Implementation
|
22
|
+
#
|
23
|
+
# DumbNumbSet is backed by a Hash of integers. Each key represents a multiple
|
24
|
+
# of the native integer length, and each value is a bit field. Each bit in the
|
25
|
+
# value represents the presence or absence of an integer.
|
26
|
+
#
|
27
|
+
# #### Performance
|
28
|
+
#
|
29
|
+
# Performance is nearly the same as a Hash, except when inserting many
|
30
|
+
# non-consecutive values.
|
31
|
+
#
|
32
|
+
# #### Size
|
33
|
+
#
|
34
|
+
# For consecutive values, DumbNumbSet is typically ~95% smaller than a Hash when
|
35
|
+
# comparing serialized size with `Marshal.dump`.
|
36
|
+
#
|
37
|
+
class DumbNumbSet
|
38
|
+
def initialize
|
39
|
+
@bitsets = {}
|
40
|
+
|
41
|
+
# Set divisor so that bit-wise operations are always performed
|
42
|
+
# with Fixnums.
|
43
|
+
if 1.size == 4
|
44
|
+
@div = 29
|
45
|
+
else
|
46
|
+
@div = 61
|
47
|
+
end
|
48
|
+
end
|
49
|
+
|
50
|
+
# Add a non-negative integer to the set.
|
51
|
+
# Raises an ArgumentError if the number given is not a non-negative integer.
|
52
|
+
def add num
|
53
|
+
check_num num
|
54
|
+
|
55
|
+
index = bitset_index num
|
56
|
+
|
57
|
+
bitset = @bitsets[index]
|
58
|
+
|
59
|
+
if bitset
|
60
|
+
@bitsets[index] = (bitset | bin_index(num))
|
61
|
+
else
|
62
|
+
@bitsets[index] = bin_index(num)
|
63
|
+
end
|
64
|
+
|
65
|
+
self
|
66
|
+
end
|
67
|
+
|
68
|
+
alias << add
|
69
|
+
|
70
|
+
# Remove number from set.
|
71
|
+
# Raises an ArgumentError if the number given is not a non-negative integer.
|
72
|
+
def remove num
|
73
|
+
check_num num
|
74
|
+
|
75
|
+
index = bitset_index num
|
76
|
+
|
77
|
+
bitset = @bitsets[index]
|
78
|
+
|
79
|
+
return false unless bitset
|
80
|
+
|
81
|
+
@bitsets[index] = (bitset ^ bin_index(num))
|
82
|
+
|
83
|
+
if @bitsets[index] == 0
|
84
|
+
@bitsets.delete index
|
85
|
+
end
|
86
|
+
|
87
|
+
num
|
88
|
+
end
|
89
|
+
|
90
|
+
# Returns true if the given number is in the set.
|
91
|
+
# Raises an ArgumentError if the number given is not a non-negative integer.
|
92
|
+
def include? num
|
93
|
+
check_num num
|
94
|
+
|
95
|
+
index = bitset_index num
|
96
|
+
|
97
|
+
bitset = @bitsets[index]
|
98
|
+
|
99
|
+
return false unless bitset
|
100
|
+
|
101
|
+
bitset & bin_index(num) != 0
|
102
|
+
end
|
103
|
+
|
104
|
+
# Returns number of keys in set (not number of values).
|
105
|
+
def size
|
106
|
+
@bitsets.length
|
107
|
+
end
|
108
|
+
|
109
|
+
private
|
110
|
+
|
111
|
+
# Calculate the bitmask to index the given number in the bitset.
|
112
|
+
def bin_index num
|
113
|
+
1 << (num % @div)
|
114
|
+
end
|
115
|
+
|
116
|
+
# Calculate the key of the bitset for a given number.
|
117
|
+
def bitset_index num
|
118
|
+
(num / @div)
|
119
|
+
end
|
120
|
+
|
121
|
+
# Check that the argument is a valid number.
|
122
|
+
def check_num num
|
123
|
+
unless num.is_a? Fixnum and num.integer? and num >= 0
|
124
|
+
raise ArgumentError, "Argument must be positive integer"
|
125
|
+
end
|
126
|
+
end
|
127
|
+
end
|
metadata
ADDED
@@ -0,0 +1,54 @@
|
|
1
|
+
--- !ruby/object:Gem::Specification
|
2
|
+
name: dumb_numb_set
|
3
|
+
version: !ruby/object:Gem::Version
|
4
|
+
version: 0.0.1
|
5
|
+
prerelease:
|
6
|
+
platform: ruby
|
7
|
+
authors:
|
8
|
+
- Justin Collins
|
9
|
+
autorequire:
|
10
|
+
bindir: bin
|
11
|
+
cert_chain: []
|
12
|
+
date: 2013-06-13 00:00:00.000000000 Z
|
13
|
+
dependencies: []
|
14
|
+
description: ! 'Ever needed to compactly store and query a set of mostly consecutive,
|
15
|
+
|
16
|
+
non-negative integers? Probably not, but if you do this library may work for
|
17
|
+
|
18
|
+
you. It''s just about as fast as a Set and a lot smaller for numbers that stay
|
19
|
+
|
20
|
+
close together.
|
21
|
+
|
22
|
+
'
|
23
|
+
email:
|
24
|
+
executables: []
|
25
|
+
extensions: []
|
26
|
+
extra_rdoc_files: []
|
27
|
+
files:
|
28
|
+
- lib/dumb_numb_set.rb
|
29
|
+
homepage: https://github.com/presidentbeef/dumb-numb-set
|
30
|
+
licenses:
|
31
|
+
- MIT
|
32
|
+
post_install_message:
|
33
|
+
rdoc_options: []
|
34
|
+
require_paths:
|
35
|
+
- lib
|
36
|
+
required_ruby_version: !ruby/object:Gem::Requirement
|
37
|
+
none: false
|
38
|
+
requirements:
|
39
|
+
- - ! '>='
|
40
|
+
- !ruby/object:Gem::Version
|
41
|
+
version: '0'
|
42
|
+
required_rubygems_version: !ruby/object:Gem::Requirement
|
43
|
+
none: false
|
44
|
+
requirements:
|
45
|
+
- - ! '>='
|
46
|
+
- !ruby/object:Gem::Version
|
47
|
+
version: '0'
|
48
|
+
requirements: []
|
49
|
+
rubyforge_project:
|
50
|
+
rubygems_version: 1.8.25
|
51
|
+
signing_key:
|
52
|
+
specification_version: 3
|
53
|
+
summary: A compact data structure for mostly consecutive, non-negative integers.
|
54
|
+
test_files: []
|