dwc_agent 1.4.4 → 1.4.5

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: fa4fb87ea91fd1f0e67278590192a55bfc7f1e8d6f4b8dc92c1f9f5eb508e44c
-  data.tar.gz: a89b51ea705885713ef8615c67e1ea10798abfe593b5646b4de9fb8e1b478762
+  metadata.gz: 7b8890828a04d30f7989f3545df032be216d1bab086f1ec6576cbe93569231bd
+  data.tar.gz: d94e120615540e65f29a13132d46b34639e3380e2265439fd05cf5febc580ff6
 SHA512:
-  metadata.gz: d676b64441d0097bd6272e2cd694c5754c4bdaed8fd0f523ecbe28748c8ccedffd9dd1c0430f5ad25cf48c02705b8131a9ce1021c07965da0791ae5f62e36c8a
-  data.tar.gz: 00effae1b438e6d97ef8da8383aa407985876d148b5b30a51ca98d2befa0dc8ac4a8c69bb389f75dd08c147853522490a406470dd8c4aa20d976238cf2cb4d82
+  metadata.gz: f2e30d42a8c2d743791c2322a51d3fe022e5ef458f68ebd30763cfc74e5ec9a1914faca1ff90cfb3364bead66cde22bccbc0ab47aef6a7d17a78c6954c8cbeff
+  data.tar.gz: d2de0c26c64f77fd273db67465d8772aef5480eb0c4dd1b48446dbb6a6495164bc9e2fbd133280c18ef7e975f82bf9f46dcd65e87727531fa7c514ebc09e0be7

data/lib/dwc_agent/constants.rb

@@ -15,7 +15,6 @@ module DwcAgent
     \b[,;]?\s*(?i:ann?onymous)\b|
     \b[,;]?\s*\(?(?i:undetermined|indeterminable|dummy|interim|accession|ill(eg|is)ible|scripsit)\)?\b|
     \b[,;]?\s*(?i:importer|gift)\:?\b|
-    \b[,;]?\s*(?i:frère|frere|père|pere|soeur|sister|bro)\.?(\b|\z)|
     \b[,;]?\s*(?i:string)\b|
     \b[,;]?\s*(?i:person\s*string)\b|
     \b[,;]?\s*(?i:colls)\.(\b|\z)|
@@ -154,10 +153,12 @@ module DwcAgent
   }
 
   PHRASE_SUBS = {
-    'Dr\.' => 'Dr. ',
-    'Mr\.' => 'Mr. ',
-    'Mrs\.' => 'Mrs. ',
-    'Prof\.' => 'Prof. '
+    'dr\.' => 'Dr. ',
+    'mr\.' => 'Mr. ',
+    'mrs\.' => 'Mrs. ',
+    'prof\.' => 'Prof. ',
+    '\, ph\.d\.' => ' Ph.D.',
+    '\, bro\.' => ' Bro.'
   }
 
   COMPLEX_SEPARATORS = %r{
@@ -258,6 +259,6 @@ module DwcAgent
     "has not"
   ]
 
-  TITLE = /\s*\b(sir|count(ess)?|(gen|adm|col|maj|capt|cmdr|lt|sgt|cpl|pvt|prof|dr|md|ph\.?d|rev|docteur|mme|abbé|ptre)\.?|frère|frere|père|pere|professor|esq\.?)(\s+|$)/i
+  TITLE = /\s*\b(sir|count(ess)?|(gen|adm|col|maj|capt|cmdr|lt|sgt|cpl|pvt|prof|dr|md|ph\.?d|rev|mme|abbé|ptre|bro|esq)\.?|docteur|father|cantor|vicar|père|pastor|rabbi|reverend|pere|soeur|sister|professor)(\s+|$)/i
 
 end

data/lib/dwc_agent/parser.rb

@@ -8,7 +8,7 @@ module DwcAgent
     end
 
     def initialize
-      options = { 
+      options = {
         prefer_comma_as_separator: true,
         separator: SPLIT_BY,
         title: TITLE
@@ -18,11 +18,11 @@ module DwcAgent
       @strip_out_regex = Regexp.new STRIP_OUT.to_s
       @residual_terminators_regex = Regexp.new SPLIT_BY.to_s + %r{\s*\z}.to_s
       @char_subs_regex = Regexp.new [CHAR_SUBS.keys.join].to_s
-      @phrase_subs_regex = Regexp.new (PHRASE_SUBS.keys.join('|')).to_s
+      @phrase_subs_regex = Regexp.new((PHRASE_SUBS.keys.join('|')).to_s, Regexp::IGNORECASE)
       @complex_separators_regex = Regexp.new COMPLEX_SEPARATORS.to_s
       @add_separators_regex = Regexp.new %r{(\S{1}\.)([[:alpha:]]{2,})}.to_s
     end
-      
+
     # Parses the passed-in string and returns a list of names.
     #
     # @param names [String] the name or names to be parsed
@@ -42,4 +42,4 @@ module DwcAgent
     end
 
   end
-end
+end

data/lib/dwc_agent/version.rb

@@ -3,7 +3,7 @@ module DwcAgent
 
     MAJOR = 1
     MINOR = 4
-    PATCH = 4
+    PATCH = 5
     BUILD = nil
 
     def self.version

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: dwc_agent
 version: !ruby/object:Gem::Version
-  version: 1.4.4
+  version: 1.4.5
 platform: ruby
 authors:
 - David P. Shorthouse
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2020-01-15 00:00:00.000000000 Z
+date: 2020-01-22 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: namae