dwc_agent 3.0.0.7 → 3.0.0.8
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.
- checksums.yaml +4 -4
- data/lib/dwc_agent/cleaner.rb +40 -13
- data/lib/dwc_agent/constants.rb +22 -0
- data/lib/dwc_agent/version.rb +1 -1
- metadata +2 -2
checksums.yaml
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
---
|
2
2
|
SHA256:
|
3
|
-
metadata.gz:
|
4
|
-
data.tar.gz:
|
3
|
+
metadata.gz: a2ec2d89d805d35fb64401736263556d13d244407456b909996912d4eb749653
|
4
|
+
data.tar.gz: 1bb39f6a59ae242362266f8a949626cfd1289d194e150463dd2eb650b98a4171
|
5
5
|
SHA512:
|
6
|
-
metadata.gz:
|
7
|
-
data.tar.gz:
|
6
|
+
metadata.gz: f005ba01365f8226966aae1e2e70ffcbbc3ffef391eeb319baad0e666c91c8edc0bef2acd1ce98dabf5d6503c063f43341c7da72e4f4b11e370537ec4592408e
|
7
|
+
data.tar.gz: bef79f42b26962c60357c3dba78d15d4b9f8189cc3c31f7034e65f78f6bc41c8b3153083354186869e9ac6b7285aa355a2b62b4c81060091a2ae99c08829ff2a
|
data/lib/dwc_agent/cleaner.rb
CHANGED
@@ -18,11 +18,14 @@ module DwcAgent
|
|
18
18
|
# @return Namae::Name [Object] a new Namae object
|
19
19
|
def clean(parsed_namae)
|
20
20
|
|
21
|
-
if parsed_namae.given &&
|
21
|
+
if parsed_namae.given &&
|
22
|
+
GIVEN_BLACKLIST.any?{ |s| s.casecmp(parsed_namae.given) == 0 }
|
22
23
|
return Namae::Name.new
|
23
24
|
end
|
24
25
|
|
25
|
-
if parsed_namae.family &&
|
26
|
+
if parsed_namae.family &&
|
27
|
+
parsed_namae.family.length == 3 &&
|
28
|
+
parsed_namae.family.count('.') == 1
|
26
29
|
return Namae::Name.new
|
27
30
|
end
|
28
31
|
|
@@ -30,7 +33,9 @@ module DwcAgent
|
|
30
33
|
return Namae::Name.new
|
31
34
|
end
|
32
35
|
|
33
|
-
if parsed_namae.given &&
|
36
|
+
if parsed_namae.given &&
|
37
|
+
parsed_namae.given.count('.') >= 3 &&
|
38
|
+
/\.\s*[a-zA-Z]{4,}\s+[a-zA-Z]{1,}\./.match(parsed_namae.given)
|
34
39
|
return Namae::Name.new
|
35
40
|
end
|
36
41
|
|
@@ -38,6 +43,13 @@ module DwcAgent
|
|
38
43
|
return Namae::Name.new
|
39
44
|
end
|
40
45
|
|
46
|
+
if parsed_namae.family &&
|
47
|
+
parsed_namae.family.count(".") == 1 &&
|
48
|
+
parsed_namae.family[-1] == "." &&
|
49
|
+
parsed_namae.family.length > 3
|
50
|
+
parsed_namae.family = parsed_namae.family.delete_suffix(".")
|
51
|
+
end
|
52
|
+
|
41
53
|
if parsed_namae.given &&
|
42
54
|
parsed_namae.family &&
|
43
55
|
parsed_namae.family.count(".") > 0 &&
|
@@ -59,6 +71,15 @@ module DwcAgent
|
|
59
71
|
parsed_namae.given = family
|
60
72
|
end
|
61
73
|
|
74
|
+
if !parsed_namae.given &&
|
75
|
+
parsed_namae.particle &&
|
76
|
+
parsed_namae.family &&
|
77
|
+
/^[A-Za-z]{3,}\s+(?:[A-Z]\.\s?){1,}$/.match(parsed_namae.family)
|
78
|
+
matched = /^(?<family>[A-Za-z]{3,})\s+(?<given>([A-Z]\.\s?){1,})$/.match(parsed_namae.family)
|
79
|
+
parsed_namae.family = matched[:family]
|
80
|
+
parsed_namae.given = matched[:given]
|
81
|
+
end
|
82
|
+
|
62
83
|
if parsed_namae.given &&
|
63
84
|
(parsed_namae.given == parsed_namae.given.upcase ||
|
64
85
|
parsed_namae.given == parsed_namae.given.downcase) &&
|
@@ -75,13 +96,21 @@ module DwcAgent
|
|
75
96
|
parsed_namae.given = NameCase(parsed_namae.given)
|
76
97
|
end
|
77
98
|
|
78
|
-
if parsed_namae.family &&
|
99
|
+
if parsed_namae.family &&
|
100
|
+
FAMILY_BLACKLIST.any?{ |s| s.casecmp(parsed_namae.family) == 0 }
|
79
101
|
return Namae::Name.new
|
80
102
|
end
|
81
103
|
|
104
|
+
if parsed_namae.family.nil? &&
|
105
|
+
!parsed_namae.given.nil? &&
|
106
|
+
!parsed_namae.given.include?(".")
|
107
|
+
parsed_namae.family = parsed_namae.given
|
108
|
+
parsed_namae.given = nil
|
109
|
+
end
|
110
|
+
|
82
111
|
parsed_namae.normalize_initials
|
83
112
|
|
84
|
-
family = parsed_namae.family
|
113
|
+
family = parsed_namae.family
|
85
114
|
given = parsed_namae.given.strip rescue nil
|
86
115
|
particle = parsed_namae.particle.strip rescue nil
|
87
116
|
appellation = parsed_namae.appellation.strip rescue nil
|
@@ -92,12 +121,10 @@ module DwcAgent
|
|
92
121
|
given = given.gsub(".", ". ").strip
|
93
122
|
end
|
94
123
|
|
95
|
-
if family.nil? &&
|
96
|
-
|
97
|
-
|
98
|
-
|
99
|
-
|
100
|
-
if !family.nil? && given.nil? && !particle.nil?
|
124
|
+
if !family.nil? &&
|
125
|
+
given.nil? &&
|
126
|
+
!particle.nil? &&
|
127
|
+
!PARTICLES.include?(particle.downcase)
|
101
128
|
given = particle.sub(/[a-z]\./, &:upcase).sub(/^(.)/) { $1.capitalize }
|
102
129
|
particle = nil
|
103
130
|
end
|
@@ -110,8 +137,8 @@ module DwcAgent
|
|
110
137
|
family = NameCase(family)
|
111
138
|
end
|
112
139
|
|
113
|
-
if !family.nil? && family.match(/[A-Z]$/)
|
114
|
-
|
140
|
+
if !family.nil? && family.match(/[A-Z]{1,3}$/)
|
141
|
+
family = NameCase(family.upcase)
|
115
142
|
end
|
116
143
|
|
117
144
|
if given.nil? && !family.nil? && family.match(/^[A-Z]{2}/)
|
data/lib/dwc_agent/constants.rb
CHANGED
@@ -342,4 +342,26 @@ module DwcAgent
|
|
342
342
|
|
343
343
|
SUFFIX = /\s*\b(JR|Jr|jr|SR|Sr|sr|[IVX]{2,})(\.|\b)/
|
344
344
|
|
345
|
+
PARTICLES = [
|
346
|
+
"ap",
|
347
|
+
"da",
|
348
|
+
"de",
|
349
|
+
"de'",
|
350
|
+
"del",
|
351
|
+
"der",
|
352
|
+
"des",
|
353
|
+
"di",
|
354
|
+
"do",
|
355
|
+
"dos",
|
356
|
+
"du",
|
357
|
+
"el",
|
358
|
+
"le",
|
359
|
+
"la",
|
360
|
+
"van",
|
361
|
+
"von",
|
362
|
+
"the",
|
363
|
+
"of",
|
364
|
+
"van der"
|
365
|
+
]
|
366
|
+
|
345
367
|
end
|
data/lib/dwc_agent/version.rb
CHANGED
metadata
CHANGED
@@ -1,14 +1,14 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: dwc_agent
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version: 3.0.0.
|
4
|
+
version: 3.0.0.8
|
5
5
|
platform: ruby
|
6
6
|
authors:
|
7
7
|
- David P. Shorthouse
|
8
8
|
autorequire:
|
9
9
|
bindir: bin
|
10
10
|
cert_chain: []
|
11
|
-
date: 2022-01-
|
11
|
+
date: 2022-01-21 00:00:00.000000000 Z
|
12
12
|
dependencies:
|
13
13
|
- !ruby/object:Gem::Dependency
|
14
14
|
name: namae
|