dwc_agent 3.0.7.0 → 3.0.9.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 309d97f743627171e1479527de5354f5f0f172b88e3cc24f40e0cbe8bc3d1ee1
-  data.tar.gz: 1d45999c8bf87e5ff68e3479ed24b7c9910e82f21295895fc3ab08e2f47cc301
+  metadata.gz: c0d83fd4dba1ddb6b9976bb1dea4c95a5365cf31b9c1afc335e035f70317a40f
+  data.tar.gz: 2fc6de4a6de283d9cf8d813306bb034ca67e20e684ce6136d6f2366512633d34
 SHA512:
-  metadata.gz: ed8b465cf99cf1a63d03b4bb23ef972a840d33a2c107c66f5a2993d95071ec70a2d97e7dbbecdbc3bee16491908118684b9b35260853bc83c7825e86725b2e92
-  data.tar.gz: f02efd62c4917e326ab111c45598f82b2d1c462d525a12753a88cacb43d5e6054e005b25a16b2522151dc06d08651a12deae3499c4d51e4f5a7bbbb474f78539
+  metadata.gz: d6ca8f0a7507c8a8d09b183f59ca47d97ff30a0de3e017da155e0b1e57ae5d34df4328ae9bdbb2fd4682b0d57afac7e251baebaa305fea5bc46b2886b22e5385
+  data.tar.gz: b9ad4777ade9052a3cd54926173c77973f899e76b3f65f43f31b45481de746e70daa4ecdcbb0b190567c27f3496d69cb5093d35522243b284ac5d4931fbe09a0

data/lib/dwc_agent/constants.rb

@@ -161,6 +161,10 @@ module DwcAgent
     \b(?i:via|from)\s*\b
   }x
 
+  POST_STRIP_TIDY = %r{
+    ^\s*[&,;]\s*
+  }x
+
   CHAR_SUBS = {
     '"' => '\'',
     '|' => ' | ',
@@ -199,15 +203,18 @@ module DwcAgent
     ' jr.,' => ' Jr.;',
     ' jr,' => ' Jr.;',
     '-jr' => ' Jr.',
-    '-Jr' => ' Jr.'
+    '-Jr' => ' Jr.',
+    'Dr.' => 'Dr. ',
+    'prof.' => 'Prof. '
   }
 
   SEPARATORS = {
+    "^(\\S{4,}),\\s+(Mrs?\\.|MRS?\\.)\\s+([A-Za-z\\.\\s]{1,})$" => "\\2 \\3 \\1",
     "^([A-Z]{1}\\.\\s*[[:alpha:]]{1,}),\\s*?([A-Z.]{1,})$" => "\\1 \\2",
     "^(\\S{4,},\\s+(?:\\S\\.\\s*){1,})\\s+(\\S{4,},\\s+(?:\\S\.\\s*){1,})$" => "\\1 | \\2",
     "(\\S{1}\\.)([[:alpha:]]{2,})" => "\\1 \\2",
     "^([[:alpha:]]{2,})(?:\\s+)((?:\\S{1}\\.\\s?){1,})$" => "\\1, \\2",
-    "([[:alpha:]]*),?\\s+(.*)\\s+(van|von)$" => "\\3 \\1, \\2",
+    "([[:alpha:]]*),?\\s*(.*)\\s+(van|von)$" => "\\3 \\1, \\2",
     "^([A-Z.\\s]+)\\s+(?:and|&|et|e)\\s+([A-Z.\\s]+)\\s+([[:alpha:]]{2,})\\s+([[:alpha:]]{2,})$" => "\\1 \\4 | \\2 \\3 \\4",
     "^([A-Z.\\s]+)\\s+(?:and|&|et|e)\\s+([A-Z.\\s]+)\\s+([[:alpha:]]{2,})(.*)$" => "\\1 \\3 | \\2 \\3 | \\4",
     "^([A-Z][[:alpha:]]{2,}),\\s*?([A-Z][[:alpha:]]{2,})\\s*?(?i:and|&|et|e|,)\\s+([A-Z][[:alpha:]]{2,})$" => "\\1 | \\2 | \\3",
@@ -219,7 +226,7 @@ module DwcAgent
     (?i:abundant)|
     (?i:adult|juvenile)|
     (?i:administra(d|t)or)|
-    (?i:anon)|
+    ^(?i:anon)$|
     (?i:australian?)|
     (?i:average)|
     (?i:believe|unclear|ill?egible|none|suggested|(dis)?agrees?)|approach|
@@ -329,6 +336,8 @@ module DwcAgent
     "inst",
     "nomenclatural",
     "orig",
+    "prof",
+    "professional",
     "qld",
     "registration",
     "science",
@@ -346,7 +355,7 @@ 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)\.?|doct(eu|o)r|father|cantor|vicar|père|pastor|rabbi|reverend|pere|soeur|sister|professor)(\s+|$)/i
+  TITLE = /\s*\b(sir|count(ess)?|colonel|(gen|adm|col|maj|capt|cmdr|lt|sgt|cpl|pvt|proff?|dr|dra\.|drª|md|ph\.?d|rev|mme|abbé|ptre|bro|esq)\.?|doct(eu|o)r|father|cantor|vicar|père|pastor|profa\.?|profª|rabbi|reverend|pere|soeur|sister|professor)(\s+|$)/i
 
   APPELLATION = /\s*\b((mrs?|ms|fr|hr)\.?|miss|herr|frau)(\s+|$)/i
 

data/lib/dwc_agent/parser.rb

@@ -18,6 +18,7 @@ module DwcAgent
       }
       @namae = Namae::Parser.new(options)
       @strip_out_regex = Regexp.new STRIP_OUT.to_s
+      @tidy_remains_regex = Regexp.new POST_STRIP_TIDY.to_s
       @char_subs_regex = Regexp.new [CHAR_SUBS.keys.join].to_s
       @phrase_subs_regex = Regexp.new PHRASE_SUBS.keys.map{|a| Regexp.escape a }.join('|').to_s
       @residual_terminators_regex = Regexp.new SPLIT_BY.to_s + %r{\s*\z}.to_s
@@ -31,6 +32,7 @@ module DwcAgent
     def parse(name)
       return [] if name.nil? || name == ""
       name.gsub!(@strip_out_regex, ' ')
+      name.gsub!(@tidy_remains_regex, '')
       name.gsub!(Regexp.union(@char_subs_regex, @phrase_subs_regex), CHAR_SUBS.merge(PHRASE_SUBS))
       @separators.each{|k| name.gsub!(k[0], k[1])}
       name.gsub!(@residual_terminators_regex, '')

data/lib/dwc_agent/version.rb

@@ -4,7 +4,7 @@ module DwcAgent
 
     MAJOR = 3
     MINOR = 0
-    PATCH = 7
+    PATCH = 9
     BUILD = 0
 
     def self.version

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: dwc_agent
 version: !ruby/object:Gem::Version
-  version: 3.0.7.0
+  version: 3.0.9.0
 platform: ruby
 authors:
 - David P. Shorthouse
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2023-08-02 00:00:00.000000000 Z
+date: 2023-09-11 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: namae