dwc_agent 3.0.0.2 → 3.0.0.6

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 4eef2473cbf2e8ce7d504a4ebce69aa9a3a9da5a6ca92018044a154ced201a9a
-  data.tar.gz: 4337d9f5441da1e6a1970449ef381e99027aeb1efe22fa0f43270d0f2c122d8b
+  metadata.gz: 7e33855504e8b60e240c2ccf1efb192dcbeb9878fb5d2f0172772ec2c2d3e9fa
+  data.tar.gz: 260e2c9304731ad852b0587a22938b31d48aeb1bd7d922eb30755e795ec054fe
 SHA512:
-  metadata.gz: bf4ad079202f2f5e7550b79e095649416d64d6bd74300127cf2272da0f3a11e77ab00d1b95c2867256c6f98e6247a725f9cf4261fa07db501b8c97f9adc5efaa
-  data.tar.gz: 52265090309fda51f66b54cc522f20c1f0415b451aea8749b0457360ae4079301ff6fffcca3fa54de60a0da1737158c35d96683ca04c4c2806189503402710d6
+  metadata.gz: bb32cd909dd7d824bc7e064316ce3e10ee6cc6b14a12371cc8464f94f2b5d0b2f0b605aeef924e44423f1af8c2af62c001f67511e2dd438086f097b81da29b57
+  data.tar.gz: cf52be5fde1f4c7ebee816fd07ae3903c3f7c55a6a41542fd21cb265a868590a3391c0d17a514c2b40ac7ca7dbfd37d9cb53211aec4a0eb0c1ec9793aa4b7c6d

data/lib/dwc_agent/constants.rb

@@ -56,6 +56,9 @@ module DwcAgent
     (?i:autres?\s+de|probab|likely|possibl(e|y)|doubtful)|
     \b\s*(?i:maybe)\s*\b|
     \b\s*(?i:prob)\.\s*\b|
+    \b\s*(?i:field\s*number)|
+    \b\s*?(?i:malaise|light|pitfall|pan|suction|lobster|actinic light|cdc)\s*(?i:trap)\s*\b|
+    \|\s*(?i:collector\s*(field\s*)?number).*$|
     \(?[,]?\s*?(?i:(local)?\s?collectors?|data\s*recorder|netter|(oper|prepar)ator)\(?s?\)?\.?\:?|
     \b[.-–,;:]?\s*(?i:department|faculty)\s*?(?i:of)?\s*?(?i:entomology|biology|zoology)|
     (?i:Engº|Agrº|Fcº|Drº|Mº|Profº|Dº|Fº)|
@@ -196,11 +199,11 @@ module DwcAgent
   SEPARATORS = {
     "^(\\S{4,},\\s+(?:\\S\\.\\s*){1,})\\s+(\\S{4,},\\s+(?:\\S\.\\s*){1,})$" => "\\1 | \\2",
     "(\\S{1}\\.)([[:alpha:]]{2,})" => "\\1 \\2",
-    "(\\w*),?\\s+(.*)\\s+(van|von)$" => "\\3 \\1, \\2",
-    "^([A-Z.\\s]+)\\s+(?:and|&|et|e)\\s+([A-Z.\\s]+)\\s+(\\w{2,})(.*)" => "\\1 \\3 | \\2 \\3 | \\4",
-    "^([A-Z]\\w{2,}),\\s*?([A-Z]\\w{2,})\\s*?[,&]\\s*?([A-Z]\\w{2,})$" => "\\1 | \\2 | \\3",
-    "^([A-Z]\\w{2,}),\\s*?([A-Z]\\w{2,}),\\s*?([A-Z]\\w{2,})\\s*?[,&]\\s*?([A-Z]\\w{3,})$" => "\\1 | \\2 | \\3 | \\4",
-    "^([A-Z]\\w{2,}),\\s*?([A-Z]\\w{2,}),\\s*?([A-Z]\\w{2,}),\\s*?([A-Z]\\w{2,})\\s*?[,&]\\s*?([A-Z]\\w{3,})$" => "\\1 | \\2 | \\3 | \\4 | \\5"
+    "([[:alpha:]]*),?\\s+(.*)\\s+(van|von)$" => "\\3 \\1, \\2",
+    "^([A-Z.\\s]+)\\s+(?:and|&|et|e)\\s+([A-Z.\\s]+)\\s+([[:alpha:]]{2,})(.*)" => "\\1 \\3 | \\2 \\3 | \\4",
+    "^([A-Z][[:alpha:]]{2,}),\\s*?([A-Z][[:alpha:]]{2,})\\s*?[,&]\\s*?([A-Z][[:alpha:]]{2,})$" => "\\1 | \\2 | \\3",
+    "^([A-Z][[:alpha:]]{2,}),\\s*?([A-Z][[:alpha:]]{2,}),\\s*?([A-Z][[:alpha:]]{2,})\\s*?[,&]\\s*?([A-Z][[:alpha:]]{3,})$" => "\\1 | \\2 | \\3 | \\4",
+    "^([A-Z][[:alpha:]]{2,}),\\s*?([A-Z][[:alpha:]]{2,}),\\s*?([A-Z][[:alpha:]]{2,}),\\s*?([A-Z][[:alpha:]]{2,})\\s*?[,&]\\s*?([A-Z][[:alpha:]]{3,})$" => "\\1 | \\2 | \\3 | \\4 | \\5"
   }
 
   BLACKLIST = %r{
@@ -290,16 +293,20 @@ module DwcAgent
     "von",
     "the",
     "of",
+    "new",
+    "no",
     "adjustment",
     "annotator",
     "available",
     "arachnology",
     "catalogue",
+    "comments",
     "curators",
     "data",
     "details",
     "determiner",
     "determination",
+    "dissected",
     "dissection",
     "entered",
     "erased",
@@ -329,7 +336,7 @@ module DwcAgent
     "has not"
   ]
 
-  TITLE = /\s*\b(sir|count(ess)?|colonel|(gen|adm|col|maj|capt|cmdr|lt|sgt|cpl|pvt|proff?|dr|md|ph\.?d|rev|mme|abbé|ptre|bro|esq)\.?|docteur|father|cantor|vicar|père|pastor|rabbi|reverend|pere|soeur|sister|professor)(\s+|$)/i
+  TITLE = /\s*\b(sir|count(ess)?|colonel|(gen|adm|col|maj|capt|cmdr|lt|sgt|cpl|pvt|proff?|dr|md|ph\.?d|rev|mme|abbé|ptre|bro|esq)\.?|doct(eu|o)r|father|cantor|vicar|père|pastor|rabbi|reverend|pere|soeur|sister|professor)(\s+|$)/i
 
   APPELLATION = /\s*\b((mrs?|ms|fr|hr)\.?|miss|herr|frau)(\s+|$)/i
 

data/lib/dwc_agent/version.rb

@@ -4,7 +4,7 @@ module DwcAgent
     MAJOR = 3
     MINOR = 0
     PATCH = 0
-    BUILD = 2
+    BUILD = 6
 
     def self.version
       [MAJOR, MINOR, PATCH, BUILD].compact.join('.').freeze

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: dwc_agent
 version: !ruby/object:Gem::Version
-  version: 3.0.0.2
+  version: 3.0.0.6
 platform: ruby
 authors:
 - David P. Shorthouse
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2021-08-16 00:00:00.000000000 Z
+date: 2021-10-11 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: namae