dwc_agent 3.1.3.0 → 3.1.5.0
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/constants.rb +13 -7
- 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: 62cb1702b9b52df587b9ac32df61a6d1c294398238f65def28ca28dff97188f3
|
|
4
|
+
data.tar.gz: 2ae483cff386f6d74f15d96bda2422117b93e0738e8312d1f79812205733d49f
|
|
5
5
|
SHA512:
|
|
6
|
-
metadata.gz:
|
|
7
|
-
data.tar.gz:
|
|
6
|
+
metadata.gz: 557e939feeb928d83ee99db7c19d992e135c574e23a9d530f1cf4f4b517f9355bfa1da13c4191b68cddbec7dcc2cea290e75723da3f6b85e4c8ba7407f08e99e
|
|
7
|
+
data.tar.gz: b3af497750749cc6585e96c28825baba87c94bcdccdeb3a50a4dae7bf99f7999717aa06399fc7c8d02735c6222f349f60ad29bf7aa599a429e44dee2fb8e7dd3
|
data/lib/dwc_agent/constants.rb
CHANGED
|
@@ -4,10 +4,15 @@ module DwcAgent
|
|
|
4
4
|
["'-]{2,}|
|
|
5
5
|
\-\.\s|
|
|
6
6
|
[,;]?\s*(?i:1st|2nd|3rd|[4-9]th)|
|
|
7
|
-
|
|
7
|
+
[,]?\s*?\d+\.\d+|
|
|
8
|
+
[,]?\s*\([#NnOo\.\s0-9\-]*[0-9a-z]+\)\s*\z|
|
|
9
|
+
[,]?\s+\#[0-9a-z]+\z|
|
|
10
|
+
[,]?\s*\#*\s+\d+\-(?i:[A-Z]|\d)+\-?\d*[A-Za-z]*\z|
|
|
11
|
+
\d*[A-Za-z]*\d*-\d*\z|
|
|
8
12
|
\b\d+\(?(?i:[[:alpha:]])\)?\b|
|
|
9
13
|
\b[,;]?\s*(?:et\.?\s+al|&\s+al)\.?|
|
|
10
14
|
\b[,;]?\s*(?i:etal)\.?|
|
|
15
|
+
\b[,;]?\s*(?i:et.al)\.?|
|
|
11
16
|
\b\s+(bis|ter)(\b|\z)|
|
|
12
17
|
\bu\.\s*a\.|
|
|
13
18
|
\b[,;]?\s*(?i:and|&)?\s*(?i:others|party)\s*\b|
|
|
@@ -27,6 +32,7 @@ module DwcAgent
|
|
|
27
32
|
^(?i:collection)\:?\s+|\s*(?i:collection)\s*$|
|
|
28
33
|
\b[,;]?\s*(?i:colls)\.(\b|\z)|
|
|
29
34
|
(?i:contactid)|
|
|
35
|
+
^(?i:dupl)[.,]{1,}|
|
|
30
36
|
\b[,;]?\s*(?i:stet)[,!]?\s*\d*\z|
|
|
31
37
|
[,;]?\s*\d+[-/\s+](?i:\d+|Jan|Feb|Mar|Apr|
|
|
32
38
|
May|Jun|Jul|Aug|Sept?|
|
|
@@ -107,8 +113,8 @@ module DwcAgent
|
|
|
107
113
|
(?i:non?)\s+(?i:specificato)|
|
|
108
114
|
\b[,;]\s+\d+\.?\z|
|
|
109
115
|
[!@?]|
|
|
116
|
+
\d{1,4}[\/.]?(?i:i|ii|iii|iv|v|vi|vii|viii|ix|x|xi|xii)[\/.]\d{1,4}|
|
|
110
117
|
[,]?\d+|
|
|
111
|
-
\s+\d+?(\/|\.)?(?i:i|ii|iii|iv|v|vi|vii|viii|ix|x)(\/|\.)\d+|
|
|
112
118
|
[,;]\z|
|
|
113
119
|
^\w{0,2}\z|
|
|
114
120
|
^[A-Z]{2,}\z|
|
|
@@ -118,9 +124,8 @@ module DwcAgent
|
|
|
118
124
|
([({].*?[)}])|
|
|
119
125
|
\s+\[([[:word:]]|[[:space:]]|[-\?\.]){10,}\]|
|
|
120
126
|
[\(\{][A-Za-z]{1,3}$|
|
|
121
|
-
\b(?i:leg)[
|
|
127
|
+
\b(?i:leg)[.:]?(\s|\z)|
|
|
122
128
|
(?:[Dd](ed|on))[\.:]|
|
|
123
|
-
\d*[A-Za-z]*\d*-\d*\z|
|
|
124
129
|
\s+[A-Z]*\d+\z|
|
|
125
130
|
\s+\d+[A-Za-z]+\z|
|
|
126
131
|
^[-,.\s;*\d]+\s?|
|
|
@@ -135,6 +140,7 @@ module DwcAgent
|
|
|
135
140
|
}x
|
|
136
141
|
|
|
137
142
|
SPLIT_BY = %r{
|
|
143
|
+
[;,]{2,}|
|
|
138
144
|
[–|ǀ∣｜│&+\/;:]|
|
|
139
145
|
\s+-\s+|
|
|
140
146
|
\s+a\.\s+|
|
|
@@ -149,7 +155,7 @@ module DwcAgent
|
|
|
149
155
|
\b(?i:checked?(\s+by)?)\s*\b|
|
|
150
156
|
\b(?i:det\.?(\s+by)?)\s*\b|
|
|
151
157
|
\b(?i:(donated)?\s*by)\s+|
|
|
152
|
-
\b(?i:dupl
|
|
158
|
+
\b(?i:dupl?[.,]?(\s+by)?|duplicate(\s+by)?)\s*\b|
|
|
153
159
|
\b(?i:ex\.?(\s+by)?|examined(\s+by)?)\s*\b|
|
|
154
160
|
\b(?i:in?dentified(\s+by)?)\s*\b|
|
|
155
161
|
\b(?i:in\s+coll\.?\s*\b)|
|
|
@@ -168,8 +174,8 @@ module DwcAgent
|
|
|
168
174
|
POST_STRIP_TIDY = %r{
|
|
169
175
|
^\s*[&,;.]\s*|
|
|
170
176
|
[\[\]]|
|
|
171
|
-
^[`'"
|
|
172
|
-
[`'"]$
|
|
177
|
+
^[`'".,!?]{1,}|
|
|
178
|
+
[`'",]{1,}$
|
|
173
179
|
}x
|
|
174
180
|
|
|
175
181
|
CHAR_SUBS = {
|
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: 3.1.
|
|
4
|
+
version: 3.1.5.0
|
|
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: 2024-
|
|
11
|
+
date: 2024-08-18 00:00:00.000000000 Z
|
|
12
12
|
dependencies:
|
|
13
13
|
- !ruby/object:Gem::Dependency
|
|
14
14
|
name: namae
|