sexmachine 0.0.5 → 0.0.6
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- data/Rakefile +0 -1
- data/lib/sexmachine/detector.rb +16 -14
- data/lib/sexmachine/version.rb +1 -1
- metadata +33 -34
- data/docs/README_rdoc.html +0 -123
- data/docs/SexMachine.html +0 -163
- data/docs/SexMachine/Detector.html +0 -376
- data/docs/created.rid +0 -5
- data/docs/index.html +0 -68
- data/docs/lib/sexmachine/detector_rb.html +0 -52
- data/docs/lib/sexmachine/version_rb.html +0 -52
- data/docs/lib/sexmachine_rb.html +0 -56
- data/docs/rdoc.css +0 -706
data/Rakefile
CHANGED
data/lib/sexmachine/detector.rb
CHANGED
|
@@ -12,12 +12,14 @@ module SexMachine
|
|
|
12
12
|
def initialize(opts = {})
|
|
13
13
|
opts = {
|
|
14
14
|
:filename => File.expand_path('../data/nam_dict.txt', __FILE__),
|
|
15
|
-
:case_sensitive => true
|
|
15
|
+
:case_sensitive => true,
|
|
16
|
+
:unknown_value => :andy
|
|
16
17
|
}.merge(opts)
|
|
17
18
|
@case_sensitive = opts[:case_sensitive]
|
|
19
|
+
@unknown_value = opts[:unknown_value]
|
|
18
20
|
parse opts[:filename]
|
|
19
21
|
end
|
|
20
|
-
|
|
22
|
+
|
|
21
23
|
def parse(fname)
|
|
22
24
|
@names = {}
|
|
23
25
|
open(fname, "r:iso8859-1:utf-8") { |f|
|
|
@@ -26,12 +28,12 @@ module SexMachine
|
|
|
26
28
|
}
|
|
27
29
|
}
|
|
28
30
|
end
|
|
29
|
-
|
|
31
|
+
|
|
30
32
|
def get_gender(name, country = nil)
|
|
31
33
|
name = UnicodeUtils.downcase(name) unless @case_sensitive
|
|
32
34
|
|
|
33
35
|
if not @names.has_key?(name)
|
|
34
|
-
|
|
36
|
+
@unknown_value
|
|
35
37
|
elsif country.nil?
|
|
36
38
|
most_popular_gender(name) { |country_values|
|
|
37
39
|
country_values.split("").select { |l| l.strip != "" }.length
|
|
@@ -45,28 +47,28 @@ module SexMachine
|
|
|
45
47
|
raise "No such country: #{country}"
|
|
46
48
|
end
|
|
47
49
|
end
|
|
48
|
-
|
|
50
|
+
|
|
49
51
|
private
|
|
50
52
|
def eat_name_line(line)
|
|
51
53
|
return if line.start_with?("#") or line.start_with?("=")
|
|
52
|
-
|
|
54
|
+
|
|
53
55
|
parts = line.split(" ").select { |p| p.strip != "" }
|
|
54
56
|
country_values = line.slice(30, line.length)
|
|
55
57
|
name = @case_sensitive ? parts[1] : UnicodeUtils.downcase(parts[1])
|
|
56
58
|
|
|
57
59
|
case parts[0]
|
|
58
|
-
|
|
59
|
-
|
|
60
|
-
|
|
61
|
-
|
|
62
|
-
|
|
63
|
-
|
|
60
|
+
when "M" then set(name, :male, country_values)
|
|
61
|
+
when "1M", "?M" then set(name, :mostly_male, country_values)
|
|
62
|
+
when "F" then set(name, :female, country_values)
|
|
63
|
+
when "1F", "?F" then set(name, :mostly_female, country_values)
|
|
64
|
+
when "?" then set(name, :andy, country_values)
|
|
65
|
+
else raise "Not sure what to do with a sex of #{parts[0]}"
|
|
64
66
|
end
|
|
65
67
|
end
|
|
66
68
|
|
|
67
69
|
def most_popular_gender(name)
|
|
68
|
-
return
|
|
69
|
-
|
|
70
|
+
return @unknown_value unless @names.has_key?(name)
|
|
71
|
+
|
|
70
72
|
max = 0
|
|
71
73
|
best = @names[name].keys.first
|
|
72
74
|
@names[name].each { |gender, country_values|
|
data/lib/sexmachine/version.rb
CHANGED
metadata
CHANGED
|
@@ -1,33 +1,37 @@
|
|
|
1
|
-
--- !ruby/object:Gem::Specification
|
|
1
|
+
--- !ruby/object:Gem::Specification
|
|
2
2
|
name: sexmachine
|
|
3
|
-
version: !ruby/object:Gem::Version
|
|
4
|
-
version: 0.0.5
|
|
3
|
+
version: !ruby/object:Gem::Version
|
|
5
4
|
prerelease:
|
|
5
|
+
version: 0.0.6
|
|
6
6
|
platform: ruby
|
|
7
|
-
authors:
|
|
7
|
+
authors:
|
|
8
8
|
- Brian Muller
|
|
9
9
|
autorequire:
|
|
10
10
|
bindir: bin
|
|
11
11
|
cert_chain: []
|
|
12
|
-
|
|
13
|
-
|
|
14
|
-
|
|
12
|
+
|
|
13
|
+
date: 2012-11-09 00:00:00 Z
|
|
14
|
+
dependencies:
|
|
15
|
+
- !ruby/object:Gem::Dependency
|
|
15
16
|
name: unicode_utils
|
|
16
|
-
|
|
17
|
+
prerelease: false
|
|
18
|
+
requirement: &id001 !ruby/object:Gem::Requirement
|
|
17
19
|
none: false
|
|
18
|
-
requirements:
|
|
19
|
-
- -
|
|
20
|
-
- !ruby/object:Gem::Version
|
|
20
|
+
requirements:
|
|
21
|
+
- - ">="
|
|
22
|
+
- !ruby/object:Gem::Version
|
|
21
23
|
version: 1.3.0
|
|
22
24
|
type: :runtime
|
|
23
|
-
|
|
24
|
-
version_requirements: *70220187561160
|
|
25
|
+
version_requirements: *id001
|
|
25
26
|
description: Get gender from first name.
|
|
26
27
|
email: brian.muller@livingsocial.com
|
|
27
28
|
executables: []
|
|
29
|
+
|
|
28
30
|
extensions: []
|
|
31
|
+
|
|
29
32
|
extra_rdoc_files: []
|
|
30
|
-
|
|
33
|
+
|
|
34
|
+
files:
|
|
31
35
|
- lib/sexmachine/data/nam_dict.txt
|
|
32
36
|
- lib/sexmachine/detector.rb
|
|
33
37
|
- lib/sexmachine/version.rb
|
|
@@ -37,37 +41,32 @@ files:
|
|
|
37
41
|
- LICENSE
|
|
38
42
|
- Rakefile
|
|
39
43
|
- README.rdoc
|
|
40
|
-
- docs/created.rid
|
|
41
|
-
- docs/index.html
|
|
42
|
-
- docs/lib/sexmachine/detector_rb.html
|
|
43
|
-
- docs/lib/sexmachine/version_rb.html
|
|
44
|
-
- docs/lib/sexmachine_rb.html
|
|
45
|
-
- docs/rdoc.css
|
|
46
|
-
- docs/README_rdoc.html
|
|
47
|
-
- docs/SexMachine/Detector.html
|
|
48
|
-
- docs/SexMachine.html
|
|
49
44
|
homepage: https://github.com/bmuller/sexmachine
|
|
50
45
|
licenses: []
|
|
46
|
+
|
|
51
47
|
post_install_message:
|
|
52
48
|
rdoc_options: []
|
|
53
|
-
|
|
49
|
+
|
|
50
|
+
require_paths:
|
|
54
51
|
- lib
|
|
55
|
-
required_ruby_version: !ruby/object:Gem::Requirement
|
|
52
|
+
required_ruby_version: !ruby/object:Gem::Requirement
|
|
56
53
|
none: false
|
|
57
|
-
requirements:
|
|
58
|
-
- -
|
|
59
|
-
- !ruby/object:Gem::Version
|
|
54
|
+
requirements:
|
|
55
|
+
- - ">="
|
|
56
|
+
- !ruby/object:Gem::Version
|
|
60
57
|
version: 1.9.0
|
|
61
|
-
required_rubygems_version: !ruby/object:Gem::Requirement
|
|
58
|
+
required_rubygems_version: !ruby/object:Gem::Requirement
|
|
62
59
|
none: false
|
|
63
|
-
requirements:
|
|
64
|
-
- -
|
|
65
|
-
- !ruby/object:Gem::Version
|
|
66
|
-
version:
|
|
60
|
+
requirements:
|
|
61
|
+
- - ">="
|
|
62
|
+
- !ruby/object:Gem::Version
|
|
63
|
+
version: "0"
|
|
67
64
|
requirements: []
|
|
65
|
+
|
|
68
66
|
rubyforge_project: sexmachine
|
|
69
|
-
rubygems_version: 1.8.
|
|
67
|
+
rubygems_version: 1.8.24
|
|
70
68
|
signing_key:
|
|
71
69
|
specification_version: 3
|
|
72
70
|
summary: Get gender from first name.
|
|
73
71
|
test_files: []
|
|
72
|
+
|
data/docs/README_rdoc.html
DELETED
|
@@ -1,123 +0,0 @@
|
|
|
1
|
-
<?xml version="1.0" encoding="utf-8"?>
|
|
2
|
-
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"
|
|
3
|
-
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
|
|
4
|
-
|
|
5
|
-
<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en">
|
|
6
|
-
<head>
|
|
7
|
-
<meta content="text/html; charset=utf-8" http-equiv="Content-Type" />
|
|
8
|
-
|
|
9
|
-
<title>File: README.rdoc [SexMachine - Gender from first name]</title>
|
|
10
|
-
|
|
11
|
-
<link type="text/css" media="screen" href="./rdoc.css" rel="stylesheet" />
|
|
12
|
-
|
|
13
|
-
<script src="./js/jquery.js" type="text/javascript"
|
|
14
|
-
charset="utf-8"></script>
|
|
15
|
-
<script src="./js/thickbox-compressed.js" type="text/javascript"
|
|
16
|
-
charset="utf-8"></script>
|
|
17
|
-
<script src="./js/quicksearch.js" type="text/javascript"
|
|
18
|
-
charset="utf-8"></script>
|
|
19
|
-
<script src="./js/darkfish.js" type="text/javascript"
|
|
20
|
-
charset="utf-8"></script>
|
|
21
|
-
</head>
|
|
22
|
-
|
|
23
|
-
<body class="file">
|
|
24
|
-
<div id="metadata">
|
|
25
|
-
<div id="home-metadata">
|
|
26
|
-
<div id="home-section" class="section">
|
|
27
|
-
<h3 class="section-header">
|
|
28
|
-
<a href="./index.html">Home</a>
|
|
29
|
-
<a href="./index.html#classes">Classes</a>
|
|
30
|
-
<a href="./index.html#methods">Methods</a>
|
|
31
|
-
</h3>
|
|
32
|
-
</div>
|
|
33
|
-
</div>
|
|
34
|
-
|
|
35
|
-
<div id="project-metadata">
|
|
36
|
-
|
|
37
|
-
|
|
38
|
-
<div id="fileindex-section" class="section project-section">
|
|
39
|
-
<h3 class="section-header">Files</h3>
|
|
40
|
-
<ul>
|
|
41
|
-
|
|
42
|
-
<li class="file"><a href="./README_rdoc.html">README.rdoc</a></li>
|
|
43
|
-
|
|
44
|
-
</ul>
|
|
45
|
-
</div>
|
|
46
|
-
|
|
47
|
-
|
|
48
|
-
<div id="classindex-section" class="section project-section">
|
|
49
|
-
<h3 class="section-header">Class Index
|
|
50
|
-
<span class="search-toggle"><img src="./images/find.png"
|
|
51
|
-
height="16" width="16" alt="[+]"
|
|
52
|
-
title="show/hide quicksearch" /></span></h3>
|
|
53
|
-
<form action="#" method="get" accept-charset="utf-8" class="initially-hidden">
|
|
54
|
-
<fieldset>
|
|
55
|
-
<legend>Quicksearch</legend>
|
|
56
|
-
<input type="text" name="quicksearch" value=""
|
|
57
|
-
class="quicksearch-field" />
|
|
58
|
-
</fieldset>
|
|
59
|
-
</form>
|
|
60
|
-
|
|
61
|
-
<ul class="link-list">
|
|
62
|
-
|
|
63
|
-
<li><a href="./SexMachine.html">SexMachine</a></li>
|
|
64
|
-
|
|
65
|
-
<li><a href="./SexMachine/Detector.html">SexMachine::Detector</a></li>
|
|
66
|
-
|
|
67
|
-
</ul>
|
|
68
|
-
<div id="no-class-search-results" style="display: none;">No matching classes.</div>
|
|
69
|
-
</div>
|
|
70
|
-
|
|
71
|
-
|
|
72
|
-
</div>
|
|
73
|
-
</div>
|
|
74
|
-
|
|
75
|
-
<div id="documentation">
|
|
76
|
-
<h1>Sex Machine</h1>
|
|
77
|
-
<p>
|
|
78
|
-
This gem uses the underlying data from the program “gender” by
|
|
79
|
-
Jorg Michael (described <a
|
|
80
|
-
href="http://www.autohotkey.com/community/viewtopic.php?t=22000">here</a>).
|
|
81
|
-
It’s use is pretty straightforward:
|
|
82
|
-
</p>
|
|
83
|
-
<pre>
|
|
84
|
-
>> require 'sexmachine'
|
|
85
|
-
>> d = Detector.new
|
|
86
|
-
>> d.get_gender("Bob")
|
|
87
|
-
:male
|
|
88
|
-
>> d.get_gender("Sally")
|
|
89
|
-
:female
|
|
90
|
-
>> d.get_gender("Pauley") # should be androgynous
|
|
91
|
-
:andy
|
|
92
|
-
</pre>
|
|
93
|
-
<p>
|
|
94
|
-
I18N is fully supported:
|
|
95
|
-
</p>
|
|
96
|
-
<pre>
|
|
97
|
-
>> d.get_gender("�lfr�n")
|
|
98
|
-
:female
|
|
99
|
-
</pre>
|
|
100
|
-
<p>
|
|
101
|
-
If you have an alterative data file, you can pass that in as an optional
|
|
102
|
-
argument to the Detector.
|
|
103
|
-
</p>
|
|
104
|
-
<p>
|
|
105
|
-
Try to avoid creating many Detectors, as each creation means reading in the
|
|
106
|
-
data file.
|
|
107
|
-
</p>
|
|
108
|
-
<h1>Licenses</h1>
|
|
109
|
-
<p>
|
|
110
|
-
The genderator code is distributed under the GPLv3. The data file
|
|
111
|
-
nam_dict.txt is released under the GNU Free Documentation License.
|
|
112
|
-
</p>
|
|
113
|
-
|
|
114
|
-
</div>
|
|
115
|
-
|
|
116
|
-
<div id="validator-badges">
|
|
117
|
-
<p><small><a href="http://validator.w3.org/check/referer">[Validate]</a></small></p>
|
|
118
|
-
<p><small>Generated with the <a href="http://deveiate.org/projects/Darkfish-Rdoc/">Darkfish
|
|
119
|
-
Rdoc Generator</a> 1.1.6</small>.</p>
|
|
120
|
-
</div>
|
|
121
|
-
</body>
|
|
122
|
-
</html>
|
|
123
|
-
|
data/docs/SexMachine.html
DELETED
|
@@ -1,163 +0,0 @@
|
|
|
1
|
-
<?xml version="1.0" encoding="utf-8"?>
|
|
2
|
-
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"
|
|
3
|
-
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
|
|
4
|
-
<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en">
|
|
5
|
-
<head>
|
|
6
|
-
<meta content="text/html; charset=utf-8" http-equiv="Content-Type" />
|
|
7
|
-
|
|
8
|
-
<title>Module: SexMachine</title>
|
|
9
|
-
|
|
10
|
-
<link rel="stylesheet" href="./rdoc.css" type="text/css" media="screen" />
|
|
11
|
-
|
|
12
|
-
<script src="./js/jquery.js" type="text/javascript"
|
|
13
|
-
charset="utf-8"></script>
|
|
14
|
-
<script src="./js/thickbox-compressed.js" type="text/javascript"
|
|
15
|
-
charset="utf-8"></script>
|
|
16
|
-
<script src="./js/quicksearch.js" type="text/javascript"
|
|
17
|
-
charset="utf-8"></script>
|
|
18
|
-
<script src="./js/darkfish.js" type="text/javascript"
|
|
19
|
-
charset="utf-8"></script>
|
|
20
|
-
|
|
21
|
-
</head>
|
|
22
|
-
<body class="module">
|
|
23
|
-
|
|
24
|
-
<div id="metadata">
|
|
25
|
-
<div id="home-metadata">
|
|
26
|
-
<div id="home-section" class="section">
|
|
27
|
-
<h3 class="section-header">
|
|
28
|
-
<a href="./index.html">Home</a>
|
|
29
|
-
<a href="./index.html#classes">Classes</a>
|
|
30
|
-
<a href="./index.html#methods">Methods</a>
|
|
31
|
-
</h3>
|
|
32
|
-
</div>
|
|
33
|
-
</div>
|
|
34
|
-
|
|
35
|
-
<div id="file-metadata">
|
|
36
|
-
<div id="file-list-section" class="section">
|
|
37
|
-
<h3 class="section-header">In Files</h3>
|
|
38
|
-
<div class="section-body">
|
|
39
|
-
<ul>
|
|
40
|
-
|
|
41
|
-
<li><a href="./lib/sexmachine/detector_rb.html?TB_iframe=true&height=550&width=785"
|
|
42
|
-
class="thickbox" title="lib/sexmachine/detector.rb">lib/sexmachine/detector.rb</a></li>
|
|
43
|
-
|
|
44
|
-
<li><a href="./lib/sexmachine/version_rb.html?TB_iframe=true&height=550&width=785"
|
|
45
|
-
class="thickbox" title="lib/sexmachine/version.rb">lib/sexmachine/version.rb</a></li>
|
|
46
|
-
|
|
47
|
-
</ul>
|
|
48
|
-
</div>
|
|
49
|
-
</div>
|
|
50
|
-
|
|
51
|
-
|
|
52
|
-
</div>
|
|
53
|
-
|
|
54
|
-
<div id="class-metadata">
|
|
55
|
-
|
|
56
|
-
<!-- Parent Class -->
|
|
57
|
-
|
|
58
|
-
|
|
59
|
-
<!-- Namespace Contents -->
|
|
60
|
-
|
|
61
|
-
<div id="namespace-list-section" class="section">
|
|
62
|
-
<h3 class="section-header">Namespace</h3>
|
|
63
|
-
<ul class="link-list">
|
|
64
|
-
|
|
65
|
-
<li><span class="type">CLASS</span> <a href="SexMachine/Detector.html">SexMachine::Detector</a></li>
|
|
66
|
-
|
|
67
|
-
</ul>
|
|
68
|
-
</div>
|
|
69
|
-
|
|
70
|
-
|
|
71
|
-
<!-- Method Quickref -->
|
|
72
|
-
|
|
73
|
-
|
|
74
|
-
<!-- Included Modules -->
|
|
75
|
-
|
|
76
|
-
</div>
|
|
77
|
-
|
|
78
|
-
<div id="project-metadata">
|
|
79
|
-
|
|
80
|
-
|
|
81
|
-
<div id="fileindex-section" class="section project-section">
|
|
82
|
-
<h3 class="section-header">Files</h3>
|
|
83
|
-
<ul>
|
|
84
|
-
|
|
85
|
-
<li class="file"><a href="./README_rdoc.html">README.rdoc</a></li>
|
|
86
|
-
|
|
87
|
-
</ul>
|
|
88
|
-
</div>
|
|
89
|
-
|
|
90
|
-
|
|
91
|
-
<div id="classindex-section" class="section project-section">
|
|
92
|
-
<h3 class="section-header">Class Index
|
|
93
|
-
<span class="search-toggle"><img src="./images/find.png"
|
|
94
|
-
height="16" width="16" alt="[+]"
|
|
95
|
-
title="show/hide quicksearch" /></span></h3>
|
|
96
|
-
<form action="#" method="get" accept-charset="utf-8" class="initially-hidden">
|
|
97
|
-
<fieldset>
|
|
98
|
-
<legend>Quicksearch</legend>
|
|
99
|
-
<input type="text" name="quicksearch" value=""
|
|
100
|
-
class="quicksearch-field" />
|
|
101
|
-
</fieldset>
|
|
102
|
-
</form>
|
|
103
|
-
|
|
104
|
-
<ul class="link-list">
|
|
105
|
-
|
|
106
|
-
<li><a href="./SexMachine.html">SexMachine</a></li>
|
|
107
|
-
|
|
108
|
-
<li><a href="./SexMachine/Detector.html">SexMachine::Detector</a></li>
|
|
109
|
-
|
|
110
|
-
</ul>
|
|
111
|
-
<div id="no-class-search-results" style="display: none;">No matching classes.</div>
|
|
112
|
-
</div>
|
|
113
|
-
|
|
114
|
-
|
|
115
|
-
</div>
|
|
116
|
-
</div>
|
|
117
|
-
|
|
118
|
-
<div id="documentation">
|
|
119
|
-
<h1 class="module">SexMachine</h1>
|
|
120
|
-
|
|
121
|
-
<div id="description">
|
|
122
|
-
|
|
123
|
-
</div>
|
|
124
|
-
|
|
125
|
-
<!-- Constants -->
|
|
126
|
-
|
|
127
|
-
<div id="constants-list" class="section">
|
|
128
|
-
<h3 class="section-header">Constants</h3>
|
|
129
|
-
<dl>
|
|
130
|
-
|
|
131
|
-
<dt><a name="VERSION">VERSION</a></dt>
|
|
132
|
-
|
|
133
|
-
<dd class="description"></dd>
|
|
134
|
-
|
|
135
|
-
|
|
136
|
-
</dl>
|
|
137
|
-
</div>
|
|
138
|
-
|
|
139
|
-
|
|
140
|
-
<!-- Attributes -->
|
|
141
|
-
|
|
142
|
-
|
|
143
|
-
<!-- Methods -->
|
|
144
|
-
|
|
145
|
-
|
|
146
|
-
</div>
|
|
147
|
-
|
|
148
|
-
|
|
149
|
-
<div id="rdoc-debugging-section-dump" class="debugging-section">
|
|
150
|
-
|
|
151
|
-
<p>Disabled; run with --debug to generate this.</p>
|
|
152
|
-
|
|
153
|
-
</div>
|
|
154
|
-
|
|
155
|
-
<div id="validator-badges">
|
|
156
|
-
<p><small><a href="http://validator.w3.org/check/referer">[Validate]</a></small></p>
|
|
157
|
-
<p><small>Generated with the <a href="http://deveiate.org/projects/Darkfish-Rdoc/">Darkfish
|
|
158
|
-
Rdoc Generator</a> 1.1.6</small>.</p>
|
|
159
|
-
</div>
|
|
160
|
-
|
|
161
|
-
</body>
|
|
162
|
-
</html>
|
|
163
|
-
|