human_name_parser 1.0.0 → 1.0.1
Sign up to get free protection for your applications and to get access to all the features.
- data/lib/human_name_parser/name.rb +31 -21
- data/lib/human_name_parser/version.rb +1 -1
- data/spec/name_spec.rb +3 -0
- metadata +47 -63
@@ -3,6 +3,7 @@ module HumanNameParser
|
|
3
3
|
PREFIXES = ['mr', 'ms', 'miss', 'mrs', 'sir', 'prof', 'professor', 'md', 'dr']
|
4
4
|
SUFFIXES = ['esq','esquire','jr','sr','2', 'i', 'ii','iii','iv', 'v', 'phd', 'md', 'do', 'dc', 'dds']
|
5
5
|
LAST_PREFIXES = ['al', 'bar','ben','bin','da','dal','de la', 'de', 'del', 'der', 'di', 'el', 'ibn', 'la', 'le', 'mc', 'san', 'st', 'ste', 'van', 'van der', 'van den', 'vel','von']
|
6
|
+
NAME_JOINERS = ['&', 'and']
|
6
7
|
|
7
8
|
attr_accessor :first, :middle, :last, :prefix, :suffix
|
8
9
|
|
@@ -43,7 +44,7 @@ module HumanNameParser
|
|
43
44
|
self.prefix = @split_name.shift
|
44
45
|
end
|
45
46
|
|
46
|
-
|
47
|
+
self
|
47
48
|
end
|
48
49
|
|
49
50
|
def parse_suffix
|
@@ -53,40 +54,43 @@ module HumanNameParser
|
|
53
54
|
end
|
54
55
|
self.suffix = self.suffix.join(' ')
|
55
56
|
|
56
|
-
|
57
|
+
self
|
57
58
|
end
|
58
59
|
|
59
60
|
def parse_last_name
|
60
|
-
|
61
|
-
|
62
|
-
|
63
|
-
while is_last_name_prefix?(@split_name.last)
|
61
|
+
if !is_name_joiner? @split_name[-2]
|
62
|
+
self.last = []
|
64
63
|
self.last.unshift @split_name.pop
|
65
|
-
end
|
66
64
|
|
67
|
-
|
65
|
+
while is_last_name_prefix?(@split_name.last)
|
66
|
+
self.last.unshift @split_name.pop
|
67
|
+
end
|
68
|
+
|
69
|
+
self.last = self.last.join(' ')
|
70
|
+
end
|
68
71
|
|
69
|
-
|
72
|
+
self
|
70
73
|
end
|
71
74
|
|
72
75
|
def parse_first_name
|
73
|
-
|
76
|
+
shift_distance = is_name_joiner?(@split_name[1]) ? 3 : 1
|
77
|
+
self.first = @split_name.shift(shift_distance).join(' ')
|
74
78
|
|
75
|
-
|
79
|
+
self
|
76
80
|
end
|
77
81
|
|
78
82
|
def parse_middle_name
|
79
83
|
# whatever's left
|
80
84
|
self.middle = @split_name.join ' '
|
81
85
|
|
82
|
-
|
86
|
+
self
|
83
87
|
end
|
84
88
|
|
85
89
|
def initials
|
86
90
|
_i = ''
|
87
|
-
_i += self.first.slice(0,1) if self.first
|
88
|
-
_i += self.middle.slice(0,1) if self.middle
|
89
|
-
_i += self.last.slice(0,1) if self.last
|
91
|
+
_i += self.first.slice(0,1) if !blank?(self.first)
|
92
|
+
_i += self.middle.slice(0,1) if !blank?(self.middle)
|
93
|
+
_i += self.last.slice(0,1) if !blank?(self.last)
|
90
94
|
_i.upcase
|
91
95
|
end
|
92
96
|
|
@@ -97,7 +101,7 @@ module HumanNameParser
|
|
97
101
|
self.middle,
|
98
102
|
self.last,
|
99
103
|
self.suffix
|
100
|
-
].reject
|
104
|
+
].reject( &method(:blank?) ).join(' ')
|
101
105
|
end
|
102
106
|
|
103
107
|
private
|
@@ -116,7 +120,7 @@ module HumanNameParser
|
|
116
120
|
@split_name = split_first_middle_last
|
117
121
|
end
|
118
122
|
|
119
|
-
|
123
|
+
self
|
120
124
|
end
|
121
125
|
|
122
126
|
# check whether each part is a suffix
|
@@ -154,14 +158,20 @@ module HumanNameParser
|
|
154
158
|
is_ix?(SUFFIXES, string)
|
155
159
|
end
|
156
160
|
|
161
|
+
def is_name_joiner?(string)
|
162
|
+
is_ix?(NAME_JOINERS, string)
|
163
|
+
end
|
164
|
+
|
157
165
|
def is_last_name_prefix?(string)
|
158
|
-
|
159
|
-
LAST_PREFIXES.any? { |p| string.downcase.match(/^#{p}$/) }
|
166
|
+
!blank?(string) && LAST_PREFIXES.include?(string.downcase)
|
160
167
|
end
|
161
168
|
|
162
169
|
def is_ix?(kind, string)
|
163
|
-
|
164
|
-
|
170
|
+
!blank?(string) && kind.any? { |k| string =~ /\A#{k}\.?\z/i }
|
171
|
+
end
|
172
|
+
|
173
|
+
def blank?(string)
|
174
|
+
string.nil? || string == ''
|
165
175
|
end
|
166
176
|
|
167
177
|
end
|
data/spec/name_spec.rb
CHANGED
@@ -1,3 +1,4 @@
|
|
1
|
+
# encoding: utf-8
|
1
2
|
require 'human_name_parser/name'
|
2
3
|
|
3
4
|
NAMES = [
|
@@ -9,6 +10,8 @@ NAMES = [
|
|
9
10
|
['Marley Mante', ['Marley', '', 'Mante', '', '', 'MM']],
|
10
11
|
['Marley Mante PhD', ['Marley', '', 'Mante', '', 'PhD', 'MM']],
|
11
12
|
['Marley Mante, esq.', ['Marley', '', 'Mante', '', 'esq.', 'MM']],
|
13
|
+
['Amy and Brian Smith', ['Amy and Brian', '', 'Smith', '', '', 'AS']],
|
14
|
+
['Amy & Brian', ['Amy & Brian', '', '', '', '', 'A']],
|
12
15
|
["Björn Charles van der O'Malley", ['Björn', 'Charles', "van der O'Malley", '', '', 'BCV']],
|
13
16
|
['Alex Rothlenson-ben-Elsburghmohampton', ['Alex', '', 'Rothlenson-ben-Elsburghmohampton', '', '', 'AR']],
|
14
17
|
['1234 Anywhere St., North Pole, SD 22323', ['', '', '', '', '', '']]
|
metadata
CHANGED
@@ -1,61 +1,55 @@
|
|
1
|
-
--- !ruby/object:Gem::Specification
|
1
|
+
--- !ruby/object:Gem::Specification
|
2
2
|
name: human_name_parser
|
3
|
-
version: !ruby/object:Gem::Version
|
4
|
-
|
3
|
+
version: !ruby/object:Gem::Version
|
4
|
+
version: 1.0.1
|
5
5
|
prerelease:
|
6
|
-
segments:
|
7
|
-
- 1
|
8
|
-
- 0
|
9
|
-
- 0
|
10
|
-
version: 1.0.0
|
11
6
|
platform: ruby
|
12
|
-
authors:
|
7
|
+
authors:
|
13
8
|
- Adam Bachman
|
14
9
|
autorequire:
|
15
10
|
bindir: bin
|
16
11
|
cert_chain: []
|
17
|
-
|
18
|
-
|
19
|
-
|
20
|
-
dependencies:
|
21
|
-
- !ruby/object:Gem::Dependency
|
12
|
+
date: 2013-05-28 00:00:00.000000000 Z
|
13
|
+
dependencies:
|
14
|
+
- !ruby/object:Gem::Dependency
|
22
15
|
name: rake
|
23
|
-
|
24
|
-
requirement: &id001 !ruby/object:Gem::Requirement
|
16
|
+
requirement: !ruby/object:Gem::Requirement
|
25
17
|
none: false
|
26
|
-
requirements:
|
27
|
-
- -
|
28
|
-
- !ruby/object:Gem::Version
|
29
|
-
|
30
|
-
segments:
|
31
|
-
- 0
|
32
|
-
version: "0"
|
18
|
+
requirements:
|
19
|
+
- - ! '>='
|
20
|
+
- !ruby/object:Gem::Version
|
21
|
+
version: '0'
|
33
22
|
type: :development
|
34
|
-
version_requirements: *id001
|
35
|
-
- !ruby/object:Gem::Dependency
|
36
|
-
name: rspec
|
37
23
|
prerelease: false
|
38
|
-
|
24
|
+
version_requirements: !ruby/object:Gem::Requirement
|
25
|
+
none: false
|
26
|
+
requirements:
|
27
|
+
- - ! '>='
|
28
|
+
- !ruby/object:Gem::Version
|
29
|
+
version: '0'
|
30
|
+
- !ruby/object:Gem::Dependency
|
31
|
+
name: rspec
|
32
|
+
requirement: !ruby/object:Gem::Requirement
|
39
33
|
none: false
|
40
|
-
requirements:
|
41
|
-
- -
|
42
|
-
- !ruby/object:Gem::Version
|
43
|
-
|
44
|
-
segments:
|
45
|
-
- 0
|
46
|
-
version: "0"
|
34
|
+
requirements:
|
35
|
+
- - ! '>='
|
36
|
+
- !ruby/object:Gem::Version
|
37
|
+
version: '0'
|
47
38
|
type: :development
|
48
|
-
|
39
|
+
prerelease: false
|
40
|
+
version_requirements: !ruby/object:Gem::Requirement
|
41
|
+
none: false
|
42
|
+
requirements:
|
43
|
+
- - ! '>='
|
44
|
+
- !ruby/object:Gem::Version
|
45
|
+
version: '0'
|
49
46
|
description: human_name_parser is intended to split names into their component parts.
|
50
|
-
email:
|
47
|
+
email:
|
51
48
|
- adam.bachman@gmail.com
|
52
49
|
executables: []
|
53
|
-
|
54
50
|
extensions: []
|
55
|
-
|
56
51
|
extra_rdoc_files: []
|
57
|
-
|
58
|
-
files:
|
52
|
+
files:
|
59
53
|
- .gitignore
|
60
54
|
- CHANGELOG.md
|
61
55
|
- Gemfile
|
@@ -71,41 +65,31 @@ files:
|
|
71
65
|
- spec/human_name_parser_spec.rb
|
72
66
|
- spec/name_spec.rb
|
73
67
|
- spec/spec_helper.rb
|
74
|
-
has_rdoc: true
|
75
68
|
homepage: https://github.com/abachman/human_name_parser
|
76
69
|
licenses: []
|
77
|
-
|
78
70
|
post_install_message:
|
79
71
|
rdoc_options: []
|
80
|
-
|
81
|
-
require_paths:
|
72
|
+
require_paths:
|
82
73
|
- lib
|
83
|
-
required_ruby_version: !ruby/object:Gem::Requirement
|
74
|
+
required_ruby_version: !ruby/object:Gem::Requirement
|
84
75
|
none: false
|
85
|
-
requirements:
|
86
|
-
- -
|
87
|
-
- !ruby/object:Gem::Version
|
88
|
-
|
89
|
-
|
90
|
-
- 0
|
91
|
-
version: "0"
|
92
|
-
required_rubygems_version: !ruby/object:Gem::Requirement
|
76
|
+
requirements:
|
77
|
+
- - ! '>='
|
78
|
+
- !ruby/object:Gem::Version
|
79
|
+
version: '0'
|
80
|
+
required_rubygems_version: !ruby/object:Gem::Requirement
|
93
81
|
none: false
|
94
|
-
requirements:
|
95
|
-
- -
|
96
|
-
- !ruby/object:Gem::Version
|
97
|
-
|
98
|
-
segments:
|
99
|
-
- 0
|
100
|
-
version: "0"
|
82
|
+
requirements:
|
83
|
+
- - ! '>='
|
84
|
+
- !ruby/object:Gem::Version
|
85
|
+
version: '0'
|
101
86
|
requirements: []
|
102
|
-
|
103
87
|
rubyforge_project:
|
104
|
-
rubygems_version: 1.
|
88
|
+
rubygems_version: 1.8.24
|
105
89
|
signing_key:
|
106
90
|
specification_version: 3
|
107
91
|
summary: Split most American names into their component parts.
|
108
|
-
test_files:
|
92
|
+
test_files:
|
109
93
|
- spec/human_name_parser_spec.rb
|
110
94
|
- spec/name_spec.rb
|
111
95
|
- spec/spec_helper.rb
|