persistent_memoize 0.0.1
Sign up to get free protection for your applications and to get access to all the features.
- data/.gitignore +17 -0
- data/Gemfile +4 -0
- data/LICENSE.txt +23 -0
- data/README.md +133 -0
- data/Rakefile +8 -0
- data/examples/example_fibonacci.rb +41 -0
- data/lib/persistent_memoize.rb +26 -0
- data/lib/persistent_memoize/version.rb +3 -0
- data/persistent_memoize.gemspec +35 -0
- data/test/test_memoize.rb +54 -0
- metadata +108 -0
data/.gitignore
ADDED
data/Gemfile
ADDED
data/LICENSE.txt
ADDED
@@ -0,0 +1,23 @@
|
|
1
|
+
Copyright (c) 2013 Neil Kandalgaonkar
|
2
|
+
|
3
|
+
MIT License
|
4
|
+
|
5
|
+
Permission is hereby granted, free of charge, to any person obtaining
|
6
|
+
a copy of this software and associated documentation files (the
|
7
|
+
"Software"), to deal in the Software without restriction, including
|
8
|
+
without limitation the rights to use, copy, modify, merge, publish,
|
9
|
+
distribute, sublicense, and/or sell copies of the Software, and to
|
10
|
+
permit persons to whom the Software is furnished to do so, subject
|
11
|
+
to the following conditions:
|
12
|
+
|
13
|
+
The above copyright notice and this permission notice shall be
|
14
|
+
included in all copies or substantial portions of the Software.
|
15
|
+
|
16
|
+
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
|
17
|
+
EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
|
18
|
+
MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
|
19
|
+
IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR
|
20
|
+
ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF
|
21
|
+
CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
|
22
|
+
WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
|
23
|
+
|
data/README.md
ADDED
@@ -0,0 +1,133 @@
|
|
1
|
+
# PersistentMemoize
|
2
|
+
|
3
|
+
## Description
|
4
|
+
Speed up methods at the cost of disk space. Keep caches on disk to speed
|
5
|
+
up later invocations of your program.
|
6
|
+
|
7
|
+
## Installation
|
8
|
+
Add this line to your application's Gemfile:
|
9
|
+
|
10
|
+
gem 'persistent_memoize'
|
11
|
+
|
12
|
+
And then execute:
|
13
|
+
|
14
|
+
$ bundle
|
15
|
+
|
16
|
+
Or install it yourself as:
|
17
|
+
|
18
|
+
$ gem install persistent_memoize
|
19
|
+
|
20
|
+
## Synopsis
|
21
|
+
|
22
|
+
Let's imagine we have a script, _fib.rb_:
|
23
|
+
|
24
|
+
def fib(n)
|
25
|
+
return n if n < 2
|
26
|
+
fib(n-1) + fib(n-2)
|
27
|
+
end
|
28
|
+
|
29
|
+
puts fib(40)
|
30
|
+
|
31
|
+
Executing it is slow - on my machine, this is 20 seconds!
|
32
|
+
|
33
|
+
$ time ruby fib.rb
|
34
|
+
102334155
|
35
|
+
|
36
|
+
real 0m20.107s
|
37
|
+
user 0m20.067s
|
38
|
+
sys 0m0.034s
|
39
|
+
|
40
|
+
So let's add memoization.
|
41
|
+
|
42
|
+
require 'persistent_memoize'
|
43
|
+
include PersistentMemoize
|
44
|
+
|
45
|
+
def fib(n)
|
46
|
+
return n if n < 2
|
47
|
+
fib(n-1) + fib(n-2)
|
48
|
+
end
|
49
|
+
|
50
|
+
memoize(:fib, './fib-cache')
|
51
|
+
puts fib(40)
|
52
|
+
|
53
|
+
And executing it now takes about a tenth of a second.
|
54
|
+
|
55
|
+
$ time ruby ./fib.rb
|
56
|
+
102334155
|
57
|
+
|
58
|
+
real 0m0.102s
|
59
|
+
user 0m0.062s
|
60
|
+
sys 0m0.036s
|
61
|
+
|
62
|
+
What's more, it stays fast - subsequent invocations after the first one are even faster.
|
63
|
+
|
64
|
+
$ time ruby ./fib.rb
|
65
|
+
102334155
|
66
|
+
|
67
|
+
real 0m0.092s
|
68
|
+
user 0m0.060s
|
69
|
+
sys 0m0.029s
|
70
|
+
|
71
|
+
## Motivation
|
72
|
+
|
73
|
+
This library is most useful when you have a program that runs repeatedly,
|
74
|
+
but which has to incorporate calculations or API results that rarely
|
75
|
+
change.
|
76
|
+
|
77
|
+
I use it to rebuild a static blog, which is based on my postings
|
78
|
+
on other sites, fetched via API. If I tweak the look of the blog, I don't want to have to
|
79
|
+
fetch all those postings all over again. So it's a quick modification
|
80
|
+
to the library that looks up those results, to cache them locally, via
|
81
|
+
memoization. If I know that those blog postings have changed, then I
|
82
|
+
delete the cache, manually, and then it regenerates from scratch.
|
83
|
+
|
84
|
+
## Caveats
|
85
|
+
|
86
|
+
As with all memoization, the method memoized must be _pure_, that
|
87
|
+
is, the result depends solely on its arguments. Memoizing a function which
|
88
|
+
returns the current date will give you the wrong answer tomorrow.
|
89
|
+
|
90
|
+
This can't magically make every method faster; it's all trade-offs. The
|
91
|
+
thing-to-be-memoized should be more expensive than computing a hash, and
|
92
|
+
deserializing data from disk. Otherwise memoizing will make it slower.
|
93
|
+
|
94
|
+
This depends on arguments having a unique serialization via the Marshal
|
95
|
+
library. Certain Ruby constructs cannot be serialized in this way and
|
96
|
+
may raise TypeErrors. See the
|
97
|
+
[Marshal documentation](http://www.ruby-doc.org/core-2.0/Marshal.html) for details.
|
98
|
+
|
99
|
+
|
100
|
+
## Constants
|
101
|
+
PersistentMemoize::PERSISTENT_MEMOIZE_VERSION
|
102
|
+
|
103
|
+
Returns the version of this package as a String.
|
104
|
+
|
105
|
+
## Methods
|
106
|
+
PersistentMemoize#memoize(method, path)
|
107
|
+
Takes a _method_ (symbol) and caches the results of _method_, for
|
108
|
+
particular arguments, on disk, in files under _path_.
|
109
|
+
|
110
|
+
If you call _method_ again with the same arguments, _memoize_ gives
|
111
|
+
you the value from disk instead of letting the method compute the
|
112
|
+
value again.
|
113
|
+
|
114
|
+
## Acknowledgements
|
115
|
+
|
116
|
+
Based on memoize by Daniel Berg (https://github.com/djberg96/memoize)
|
117
|
+
|
118
|
+
Daniel's library has a file-oriented cache but it creates a single cache file per
|
119
|
+
method, containing all the cached results. For my workloads, this cache file
|
120
|
+
gets very large and rewriting it is slow. So persistent_memoize stores each result
|
121
|
+
in its own file.
|
122
|
+
|
123
|
+
Daniel also included this note in his code, so I might as well acknowledge
|
124
|
+
these people too:
|
125
|
+
|
126
|
+
Code borrowed from Nobu Nakada (ruby-talk:155159).
|
127
|
+
Code borrowed from Ara Howard (ruby-talk:173428).
|
128
|
+
Code borrowed from Andrew Johnson (http://tinyurl.com/8ymx8)
|
129
|
+
Ideas taken from Brian Buckley and Sean O'Halpin.
|
130
|
+
Tiny URL provided for Andrew Johnson because I could not find the ruby-talk
|
131
|
+
reference. The gateway may have been broken at the time.
|
132
|
+
|
133
|
+
|
data/Rakefile
ADDED
@@ -0,0 +1,41 @@
|
|
1
|
+
###################################################################
|
2
|
+
# fibonacci.rb
|
3
|
+
#
|
4
|
+
# Demonstrates, via Benchmark, the difference between a memoized
|
5
|
+
# version of the fibonnaci method versus a non-memoized version.
|
6
|
+
#
|
7
|
+
# You can run this via the 'example_fib' rake task.
|
8
|
+
###################################################################
|
9
|
+
require 'benchmark'
|
10
|
+
require 'fileutils'
|
11
|
+
require 'persistent_memoize'
|
12
|
+
include PersistentMemoize
|
13
|
+
|
14
|
+
# Our fibonacci function
|
15
|
+
def fib(n)
|
16
|
+
return n if n < 2
|
17
|
+
fib(n-1) + fib(n-2)
|
18
|
+
end
|
19
|
+
|
20
|
+
file = File.join((ENV['HOME'] || ENV['USERPROFILE']), 'fib.cache')
|
21
|
+
|
22
|
+
max_iter = ARGV[0].to_i
|
23
|
+
max_fib = ARGV[1].to_i
|
24
|
+
|
25
|
+
max_iter = 100 if max_iter == 0
|
26
|
+
max_fib = 25 if max_fib == 0
|
27
|
+
|
28
|
+
print "\nBenchmarking against version: " + PERSISTENT_MEMOIZE_VERSION + "\n\n"
|
29
|
+
|
30
|
+
Benchmark.bm(35) do |x|
|
31
|
+
x.report("Not memoized:"){
|
32
|
+
max_iter.times{ fib(max_fib) }
|
33
|
+
}
|
34
|
+
|
35
|
+
x.report("Memoized to file:"){
|
36
|
+
memoize(:fib, file)
|
37
|
+
max_iter.times{ fib(max_fib) }
|
38
|
+
}
|
39
|
+
end
|
40
|
+
|
41
|
+
FileUtils.remove_dir(file) if File.exists?(file)
|
@@ -0,0 +1,26 @@
|
|
1
|
+
require 'digest/md5'
|
2
|
+
require 'fileutils'
|
3
|
+
require "persistent_memoize/version"
|
4
|
+
|
5
|
+
module PersistentMemoize
|
6
|
+
# Memoize the method 'name', with results stored in files under 'path'
|
7
|
+
def memoize(name, path)
|
8
|
+
unless File.exists?(path)
|
9
|
+
FileUtils.mkdir_p(path)
|
10
|
+
end
|
11
|
+
|
12
|
+
(class<<self; self; end).send(:define_method, name) do |*args|
|
13
|
+
key = Digest::MD5.hexdigest(Marshal.dump(args))
|
14
|
+
cacheFile = File.join(path, key)
|
15
|
+
if File.exists?(cacheFile)
|
16
|
+
results = File.open(cacheFile, 'rb'){ |f| Marshal.load(f) }
|
17
|
+
else
|
18
|
+
results = super(*args)
|
19
|
+
File.open(cacheFile, 'wb'){ |f| Marshal.dump(results, f) }
|
20
|
+
end
|
21
|
+
results
|
22
|
+
end
|
23
|
+
|
24
|
+
end
|
25
|
+
|
26
|
+
end
|
@@ -0,0 +1,35 @@
|
|
1
|
+
# coding: utf-8
|
2
|
+
lib = File.expand_path('../lib', __FILE__)
|
3
|
+
$LOAD_PATH.unshift(lib) unless $LOAD_PATH.include?(lib)
|
4
|
+
require 'persistent_memoize/version'
|
5
|
+
|
6
|
+
# Build the gem with the 'rake gem:build' command.
|
7
|
+
|
8
|
+
Gem::Specification.new do |spec|
|
9
|
+
spec.name = 'persistent_memoize'
|
10
|
+
spec.version = PersistentMemoize::VERSION
|
11
|
+
spec.version = '0.0.1'
|
12
|
+
spec.authors = ['Neil Kandalgaonkar']
|
13
|
+
spec.license = 'MIT'
|
14
|
+
spec.email = ['neilk@brevity.org']
|
15
|
+
spec.homepage = 'https://github.com/neilk/persistent_memoize'
|
16
|
+
spec.platform = Gem::Platform::RUBY
|
17
|
+
spec.summary = 'Speeds up methods with caches, which persistent on disk'
|
18
|
+
spec.has_rdoc = true
|
19
|
+
spec.files = `git ls-files`.split($/)
|
20
|
+
spec.test_files = spec.files.grep(%r{^(test|spec|features)/})
|
21
|
+
spec.executables = spec.files.grep(%r{^bin/}) { |f| File.basename(f) }
|
22
|
+
spec.require_paths = ["lib"]
|
23
|
+
|
24
|
+
spec.extra_rdoc_files = ['README.md']
|
25
|
+
|
26
|
+
spec.add_development_dependency('test-unit', '>= 2.0.2')
|
27
|
+
spec.add_development_dependency "bundler", "~> 1.3"
|
28
|
+
spec.add_development_dependency "rake"
|
29
|
+
|
30
|
+
spec.description = <<-EOF
|
31
|
+
This allows you to cache method calls for faster execution, at the cost of
|
32
|
+
storage space and IO operations. The caches are kept on disk,
|
33
|
+
persistent between executions of the program.
|
34
|
+
EOF
|
35
|
+
end
|
@@ -0,0 +1,54 @@
|
|
1
|
+
###############################################
|
2
|
+
# test_memoize.rb
|
3
|
+
#
|
4
|
+
# Test suite for the memoize library.
|
5
|
+
###############################################
|
6
|
+
require 'rubygems'
|
7
|
+
gem 'test-unit'
|
8
|
+
|
9
|
+
require 'test/unit'
|
10
|
+
require 'fileutils'
|
11
|
+
require 'persistent_memoize'
|
12
|
+
|
13
|
+
class TC_PersistentMemoize < Test::Unit::TestCase
|
14
|
+
include PersistentMemoize
|
15
|
+
|
16
|
+
def setup
|
17
|
+
@path = File.join((ENV['HOME'] || ENV['USERPROFILE']), 'test.cache')
|
18
|
+
end
|
19
|
+
|
20
|
+
def add(x, y)
|
21
|
+
return x + y
|
22
|
+
end
|
23
|
+
|
24
|
+
def fib(n)
|
25
|
+
return n if n < 2
|
26
|
+
fib(n-1) + fib(n-2)
|
27
|
+
end
|
28
|
+
|
29
|
+
def factorial(n)
|
30
|
+
f = 1
|
31
|
+
n.downto(2) { |x| f *= x }
|
32
|
+
f
|
33
|
+
end
|
34
|
+
|
35
|
+
def test_memoize
|
36
|
+
assert_nothing_raised{ fib(5) }
|
37
|
+
assert_nothing_raised{ memoize(:fib, @path) }
|
38
|
+
assert_nothing_raised{ fib(50) }
|
39
|
+
assert_equal(55, fib(10))
|
40
|
+
end
|
41
|
+
|
42
|
+
def test_memoize_directory_properties
|
43
|
+
assert(!File.exists?(@path), "path did not exist")
|
44
|
+
assert_nothing_raised{ memoize(:add, @path) }
|
45
|
+
assert(File.exists?(@path), "path does exist")
|
46
|
+
assert(Dir.entries(@path).select{ |f| f != '.' and f != '..' }.length == 0, "path is empty of files")
|
47
|
+
assert_nothing_raised{ add(10, 20) }
|
48
|
+
assert(Dir.entries(@path).select{ |f| f != '.' and f != '..' }.length == 1, "path has exactly one file")
|
49
|
+
end
|
50
|
+
|
51
|
+
def teardown
|
52
|
+
FileUtils.remove_dir(@path) if File.exists?(@path)
|
53
|
+
end
|
54
|
+
end
|
metadata
ADDED
@@ -0,0 +1,108 @@
|
|
1
|
+
--- !ruby/object:Gem::Specification
|
2
|
+
name: persistent_memoize
|
3
|
+
version: !ruby/object:Gem::Version
|
4
|
+
version: 0.0.1
|
5
|
+
prerelease:
|
6
|
+
platform: ruby
|
7
|
+
authors:
|
8
|
+
- Neil Kandalgaonkar
|
9
|
+
autorequire:
|
10
|
+
bindir: bin
|
11
|
+
cert_chain: []
|
12
|
+
date: 2013-04-01 00:00:00.000000000 Z
|
13
|
+
dependencies:
|
14
|
+
- !ruby/object:Gem::Dependency
|
15
|
+
name: test-unit
|
16
|
+
requirement: !ruby/object:Gem::Requirement
|
17
|
+
none: false
|
18
|
+
requirements:
|
19
|
+
- - ! '>='
|
20
|
+
- !ruby/object:Gem::Version
|
21
|
+
version: 2.0.2
|
22
|
+
type: :development
|
23
|
+
prerelease: false
|
24
|
+
version_requirements: !ruby/object:Gem::Requirement
|
25
|
+
none: false
|
26
|
+
requirements:
|
27
|
+
- - ! '>='
|
28
|
+
- !ruby/object:Gem::Version
|
29
|
+
version: 2.0.2
|
30
|
+
- !ruby/object:Gem::Dependency
|
31
|
+
name: bundler
|
32
|
+
requirement: !ruby/object:Gem::Requirement
|
33
|
+
none: false
|
34
|
+
requirements:
|
35
|
+
- - ~>
|
36
|
+
- !ruby/object:Gem::Version
|
37
|
+
version: '1.3'
|
38
|
+
type: :development
|
39
|
+
prerelease: false
|
40
|
+
version_requirements: !ruby/object:Gem::Requirement
|
41
|
+
none: false
|
42
|
+
requirements:
|
43
|
+
- - ~>
|
44
|
+
- !ruby/object:Gem::Version
|
45
|
+
version: '1.3'
|
46
|
+
- !ruby/object:Gem::Dependency
|
47
|
+
name: rake
|
48
|
+
requirement: !ruby/object:Gem::Requirement
|
49
|
+
none: false
|
50
|
+
requirements:
|
51
|
+
- - ! '>='
|
52
|
+
- !ruby/object:Gem::Version
|
53
|
+
version: '0'
|
54
|
+
type: :development
|
55
|
+
prerelease: false
|
56
|
+
version_requirements: !ruby/object:Gem::Requirement
|
57
|
+
none: false
|
58
|
+
requirements:
|
59
|
+
- - ! '>='
|
60
|
+
- !ruby/object:Gem::Version
|
61
|
+
version: '0'
|
62
|
+
description: ! " This allows you to cache method calls for faster execution, at
|
63
|
+
the cost of \n storage space and IO operations. The caches are kept on disk,\n
|
64
|
+
\ persistent between executions of the program. \n"
|
65
|
+
email:
|
66
|
+
- neilk@brevity.org
|
67
|
+
executables: []
|
68
|
+
extensions: []
|
69
|
+
extra_rdoc_files:
|
70
|
+
- README.md
|
71
|
+
files:
|
72
|
+
- .gitignore
|
73
|
+
- Gemfile
|
74
|
+
- LICENSE.txt
|
75
|
+
- README.md
|
76
|
+
- Rakefile
|
77
|
+
- examples/example_fibonacci.rb
|
78
|
+
- lib/persistent_memoize.rb
|
79
|
+
- lib/persistent_memoize/version.rb
|
80
|
+
- persistent_memoize.gemspec
|
81
|
+
- test/test_memoize.rb
|
82
|
+
homepage: https://github.com/neilk/persistent_memoize
|
83
|
+
licenses:
|
84
|
+
- MIT
|
85
|
+
post_install_message:
|
86
|
+
rdoc_options: []
|
87
|
+
require_paths:
|
88
|
+
- lib
|
89
|
+
required_ruby_version: !ruby/object:Gem::Requirement
|
90
|
+
none: false
|
91
|
+
requirements:
|
92
|
+
- - ! '>='
|
93
|
+
- !ruby/object:Gem::Version
|
94
|
+
version: '0'
|
95
|
+
required_rubygems_version: !ruby/object:Gem::Requirement
|
96
|
+
none: false
|
97
|
+
requirements:
|
98
|
+
- - ! '>='
|
99
|
+
- !ruby/object:Gem::Version
|
100
|
+
version: '0'
|
101
|
+
requirements: []
|
102
|
+
rubyforge_project:
|
103
|
+
rubygems_version: 1.8.23
|
104
|
+
signing_key:
|
105
|
+
specification_version: 3
|
106
|
+
summary: Speeds up methods with caches, which persistent on disk
|
107
|
+
test_files:
|
108
|
+
- test/test_memoize.rb
|