dwc_agent 1.5.0.5 → 1.5.0.6

Sign up to get free protection for your applications and to get access to all the features.
checksums.yaml CHANGED
@@ -1,7 +1,7 @@
1
1
  ---
2
2
  SHA256:
3
- metadata.gz: 0445f35092b28cfbcc273d01f90eee34c9642d7fc8b9c0b7e32e7720dc2f316b
4
- data.tar.gz: f10b0b6424007a829196851ece90d6138693dcd1825b6b5712a96b9b9a2ce224
3
+ metadata.gz: 4b3360ae9a26c61f08d3a687e8c4e1af9e19334714746a976cb4c90e93d72a63
4
+ data.tar.gz: 4ada728f04c124ec1fbd6a9b45c1ecba80f45319f051a087c2074b2417110478
5
5
  SHA512:
6
- metadata.gz: 69d6e2be09b6edc55b5ebd0fef931dd5fcaf9df09b150bcd4580a9aace8c24576db65ea194352c134418f974b4cbfc0bf3254a1d4e22132f56203afcc06dbc08
7
- data.tar.gz: 467f50b23cb2d3cf8ad406b9cd14c3bdbd6635806c46304be7a0708a03755c23db56f0295dd2d89f8325dbc7ac1a3c34b237103974bb27a62ab38446838b21bd
6
+ metadata.gz: ced6e4cd9352b267ee8662993d4a39c9aa6e35965c550fef33993bd3c7c40feb07e7ad1ac8e6c60531f843b5e3d18a5bc0caafb38de41b55f0ad706be898d81a
7
+ data.tar.gz: 4c3c76a1f3cf17114d9e23a2ec19f40772fe38b5bd1b6dd49264cf82081dac797b182ae0a2cd5d90732a0f3cd947251ea28830f7769a82d4bfe8a10ca166213e
@@ -18,10 +18,6 @@ module DwcAgent
18
18
  def clean(parsed_namae)
19
19
  blank_name = { title: nil, appellation: nil, given: nil, particle: nil, family: nil, suffix: nil }
20
20
 
21
- if parsed_namae.family && FAMILY_BLACKLIST.any?{ |s| s.casecmp(parsed_namae.family) == 0 }
22
- return blank_name
23
- end
24
-
25
21
  if parsed_namae.given && GIVEN_BLACKLIST.any?{ |s| s.casecmp(parsed_namae.given) == 0 }
26
22
  return blank_name
27
23
  end
@@ -79,6 +75,10 @@ module DwcAgent
79
75
  parsed_namae.given = NameCase(parsed_namae.given)
80
76
  end
81
77
 
78
+ if parsed_namae.family && FAMILY_BLACKLIST.any?{ |s| s.casecmp(parsed_namae.family) == 0 }
79
+ return blank_name
80
+ end
81
+
82
82
  parsed_namae.normalize_initials
83
83
 
84
84
  family = parsed_namae.family.gsub(/\.\z/, '').strip rescue nil
@@ -54,10 +54,11 @@ module DwcAgent
54
54
  (?i:autres?\s+de|probab|likely|possibl(e|y)|doubtful)|
55
55
  \b\s*(?i:maybe)\s*\b|
56
56
  \b\s*(?i:prob)\.\s*\b|
57
- \(?[,]?\s*?(?i:(local)?\s?collector|data\s*recorder|netter|(oper|prepar)ator)\(?s?\)?\.?\:?|
57
+ \(?[,]?\s*?(?i:(local)?\s?collectors?|data\s*recorder|netter|(oper|prepar)ator)\(?s?\)?\.?\:?|
58
58
  \b[.-–,;:]?\s*(?i:department|faculty)\s*?(?i:of)?\s*?(?i:entomology|biology|zoology)|
59
59
  (?i:Engº|Agrº|Fcº|Drº|Mº|Profº|Dº|Fº)|
60
60
  (?i:fide)\:?\s*\b|
61
+ (?i:first\s+name\s+unknown)|
61
62
  (?i:game\s+dept)\.?\s*\b|
62
63
  (?i:see\s+notes?\s*(inside)?)|
63
64
  (?i:see\s+letter\s+enclosed)|
@@ -67,16 +68,18 @@ module DwcAgent
67
68
  \(?(?i:source)\(?|
68
69
  (?i:according\s+to)|
69
70
  (?i:lanuv)\d+|
71
+ \b\s*name\b|
72
+ \b\s*lost\b|
70
73
  (?i:nswobs)|
71
74
  ORCID|
72
75
  MRI(\s|-)PAS|
73
76
  urn\:qm\.qld\.gov\.au\:collector|
74
77
  (?i:University\s+of\s+(Southern\s+)?California(,\s+Berkeley)?)|
75
- (?i:Field\s+Museum\s+of\s+Natural\s+History)|
76
- (?i:American\s+Museum\s+of\s+Natural\s+History)|
78
+ (?i:field\s+museum\s+of\s+natural\s+history)|
79
+ (?i:american\s+museum\s+of\s+natural\s+history)|
77
80
  (?i:The\s+Paleontological\s+Research\s+Institution)|
78
81
  (?i:museums?\s+victoria)|
79
- \b\s*(?i:United\s+States|Russia)\s*\b|
82
+ \b\s*(?i:united\s+states|russia)\s*\b|
80
83
  (?i:revised|photograph|fruits\s+only)|
81
84
  -?\s*(?i:sight\s+(id|identifi?cation))\.?\s*\b|
82
85
  -?\s*(?i:synonym(y|ie))|
@@ -244,7 +247,7 @@ module DwcAgent
244
247
  (?i:though)|
245
248
  (?i:texas\s+instruments?)\s*?(for)?|
246
249
  (?:tropical)|
247
- (?i:toward|seen at)|
250
+ (?i:toward|seen\s+at)|
248
251
  (?i:unidentified|unspecified|unk?nown?|unnamed|unread|unmistak|no agent)|
249
252
  (?i:urn\:)|
250
253
  (?i:usda|ucla)|
@@ -253,30 +256,47 @@ module DwcAgent
253
256
  }x
254
257
 
255
258
  FAMILY_BLACKLIST = [
259
+ "ap",
256
260
  "da",
261
+ "de",
257
262
  "de'",
258
263
  "del",
259
264
  "der",
265
+ "di",
266
+ "do",
267
+ "dos",
260
268
  "du",
261
269
  "el",
270
+ "le",
271
+ "la",
262
272
  "van",
263
273
  "von",
264
274
  "the",
265
275
  "of",
266
276
  "adjustment",
277
+ "annotator",
267
278
  "available",
268
279
  "arachnology",
269
280
  "catalogue",
270
281
  "curators",
271
282
  "data",
283
+ "details",
284
+ "determiner",
272
285
  "determination",
273
286
  "dissection",
274
287
  "entered",
288
+ "erased",
275
289
  "indecipherable",
276
290
  "nomenclatural",
277
291
  "orig",
278
292
  "registration",
279
- "science"
293
+ "science",
294
+ "wg",
295
+ "wm",
296
+ "wn",
297
+ "zw",
298
+ "zz",
299
+ "z-"
280
300
  ]
281
301
 
282
302
  GIVEN_BLACKLIST = [
@@ -4,7 +4,7 @@ module DwcAgent
4
4
  MAJOR = 1
5
5
  MINOR = 5
6
6
  PATCH = 0
7
- BUILD = 5
7
+ BUILD = 6
8
8
 
9
9
  def self.version
10
10
  [MAJOR, MINOR, PATCH, BUILD].compact.join('.').freeze
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: 1.5.0.5
4
+ version: 1.5.0.6
5
5
  platform: ruby
6
6
  authors:
7
7
  - David P. Shorthouse
8
- autorequire:
8
+ autorequire:
9
9
  bindir: bin
10
10
  cert_chain: []
11
- date: 2020-08-18 00:00:00.000000000 Z
11
+ date: 2020-09-06 00:00:00.000000000 Z
12
12
  dependencies:
13
13
  - !ruby/object:Gem::Dependency
14
14
  name: namae
@@ -102,7 +102,7 @@ homepage: https://github.com/bionomia/dwc_agent
102
102
  licenses:
103
103
  - MIT
104
104
  metadata: {}
105
- post_install_message:
105
+ post_install_message:
106
106
  rdoc_options:
107
107
  - "--encoding"
108
108
  - UTF-8
@@ -119,8 +119,8 @@ required_rubygems_version: !ruby/object:Gem::Requirement
119
119
  - !ruby/object:Gem::Version
120
120
  version: '0'
121
121
  requirements: []
122
- rubygems_version: 3.0.6
123
- signing_key:
122
+ rubygems_version: 3.1.2
123
+ signing_key:
124
124
  specification_version: 4
125
125
  summary: Parse Darwin Core agent terms such as recordedBy and identifiedBy
126
126
  test_files: []