dwc_agent 3.4.2.0 → 3.4.3.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: df1815f4816f1738bd4b1f2860e7fbe35f247667905d0dfaa6123e32588de47f
-  data.tar.gz: 5b019b47225dcb238a1c6149b26d9d6fb2f8825e64e78f912fe4df632d98b5cc
+  metadata.gz: 29cbee65088adb6c5914a955a61f7fb6ff88e84fc72394cae0168edeb8652953
+  data.tar.gz: a79f84cd59930282009d7f86545f4ce2a726d1818693b0abd9648add0160ca41
 SHA512:
-  metadata.gz: 5cae0d260345a64d7a0c4e9ba1d59c13cddf78f3463621cfdb4fe17dd83ea6803f5eb4bddca3d021afde7f8910570b20fd77a4bef4b2113086fb39b32c29742c
-  data.tar.gz: 8599bd4e6e7083066e9af8803c98ebfe69c14258c81472f1ee1a7ec67e1f3733db7cf2a00f0b1b1460d08aa8a27a957afffe7ccbe5d182a4300321bebe477dd9
+  metadata.gz: 79b6048d32a3ccd7e8fa6c6833c5df709f7a37cd2fc3f904d4075796f669cc33fd185aedb61893648aa310bc18abad854a8b5e0687023bb9de6eae27b93aded4
+  data.tar.gz: 3ddd88b8c5d5a0b0725876e3758caffa20f6fcbf1b69a55b1967e1984f67f0d117501f84cb7cc5759e038d968df069c1d4dddff1c86a8b31afe427d3ab6c73ff

data/lib/dwc_agent/constants.rb

@@ -1,3 +1,5 @@
+require "set"
+
 module DwcAgent
   STRIP_OUT = %r{
     (?i:acc\s?\#)|
@@ -138,7 +140,7 @@ module DwcAgent
     \s+de\s*$|
     \.{2,}$|
     [^[:alnum:][:blank:][:punct:][∣´|ǀ∣｜│`~$^+|<>]]      # Removes emojis from string
-  }x
+  }x.freeze
 
   SPLIT_BY = %r{
     [;,]{2,} |                                            # Multiple semicolons or commas
@@ -170,14 +172,14 @@ module DwcAgent
     \b(?i:then(\s+by)?)\s+ |                              # "then (by)"
     \b(?i:veri?f?\.?\:?(\s+by)?|v(e|é)rifi(e|é)d?(\s+by)?)\s*\b | # "verif."
     \b(?i:via|from)\s*\b                                  # "via", "from"
-  }x
+  }x.freeze
 
   POST_STRIP_TIDY = %r{
     ^\s*[&,;.]\s* |                                       # Leading whitespace followed by any combination of &, ;, or .
     [\[\]] |                                              # Any standalone square brackets
     ^[`'".,!?]+ |                                         # Leading repeated punctuation (` ' " . , ! ?)
     [`'",]+$                                              # Trailing repeated punctuation (` ' ")
-  }x
+  }x.freeze
 
   CHAR_SUBS = {
     '"' => '\'',
@@ -219,7 +221,7 @@ module DwcAgent
     'prof.' => 'Prof. ',
     ' .;' => '. ;',
     ', &' => ' &'
-  }
+  }.freeze
 
   COMPLEX_SEPARATORS = {
     "^(\\S{4,}),\\s+(Mrs?\\.|MRS?\\.)\\s+([A-Za-z\\.\\s]+)$" => "\\2 \\3 \\1",
@@ -228,7 +230,8 @@ module DwcAgent
     "^(\\S{4,},\\s+(?:\\S\\.\\s*)+)\\s+(\\S{4,},\\s+(?:\\S\.\\s*)+)$" => "\\1 | \\2",
     "(\\S{1}\\.)([[:alpha:]]{2,})" => "\\1 \\2",
     "^([[:alpha:]]{2,})(?:\\s+)((?:\\S{1}\\.\\s?)+)$" => "\\1, \\2",
-    "([[:alpha:]]*),?\\s*(.*)\\s+(van|von|v\\.|v(a|o)n\\s+der?)$" => "\\3 \\1, \\2",
+    "^([[:alpha:]]*),?\\s*(.*)\\s+(van|von|v\\.|v(a|o)n\\s+der?)(?i:and|&|et|e|,|;)\s*([[:alpha:]]*),?\\s*(.*)\\s+(van|von|v\\.|v(a|o)n\\s+der?)$" => "\\3 \\1, \\2 | \\7 \\5, \\6",
+    "^([[:alpha:]]*),?\\s*(.*)\\s+(van|von|v\\.|v(a|o)n\\s+der?)$" => "\\3 \\1, \\2",
     "^((?i:[A-Z]\\.\\s?)+)\\s?(?:and|&|et|e)\\s+((?i:[A-Z]\\.\\s?)+)\\s+([[:alpha:]’`'-]{2,})\\s+([[:alpha:]’`'-]{2,})$" => "\\1 \\4 | \\2 \\3 \\4",
     "^((?i:[A-Z]\\.\\s?)+)\\s?(?:and|&|et|e)\\s+((?i:[A-Z]\\.\\s?)+)\\s+([[:alpha:]’`'-]{2,})(.*)$" => "\\1 \\3 | \\2 \\3 | \\4",
     "^([A-Z]{1,3})\\s+(?:and|&|et|e)\\s+([A-Z]{1,3})\\s+([[:alpha:]’`'-]{2,})(.*)$" => "\\1 \\3 | \\2 \\3 | \\4",
@@ -236,7 +239,7 @@ module DwcAgent
     "^([A-Z][[:alpha:]]{2,}),\\s*?([A-Z][[:alpha:]]{2,})\\s*?(?i:and|&|et|e|,)\\s+([A-Z][[:alpha:]]{2,})$" => "\\1 | \\2 | \\3",
     "^([A-Z][[:alpha:]]{2,}),\\s*?([A-Z][[:alpha:]]{2,}),\\s*?([A-Z][[:alpha:]]{2,})\\s*?(?i:and|&|et|e|,)\\s+([A-Z][[:alpha:]]{3,})$" => "\\1 | \\2 | \\3 | \\4",
     "^([A-Z][[:alpha:]]{2,}),\\s*?([A-Z][[:alpha:]]{2,}),\\s*?([A-Z][[:alpha:]]{2,}),\\s*?([A-Z][[:alpha:]]{2,})\\s*?(?i:and|&|et|e|,)\\s+([A-Z][[:alpha:]]{3,})$" => "\\1 | \\2 | \\3 | \\4 | \\5"
-  }
+  }.freeze
 
   BLACKLIST = %r{
     (?i:
@@ -311,9 +314,9 @@ module DwcAgent
       workshop | garden | farm | jardin | public |
       ^de$
     )
-  }x
+  }x.freeze
 
-  FAMILY_GREENLIST = [
+  FAMILY_GREENLIST = Set.new([
     "Ng",
     "Srb",
     "Srp",
@@ -323,9 +326,9 @@ module DwcAgent
     "Smrt",
     "Krc",
     "Krč"
-  ]
+  ]).freeze
 
-  FAMILY_BLACKLIST = [
+  FAMILY_BLACKLIST = Set.new([
     "a b",
     "a e",
     "a g",
@@ -390,20 +393,20 @@ module DwcAgent
     "zw",
     "zz",
     "z-"
-  ]
+  ]).freeze
 
-  GIVEN_BLACKLIST = [
+  GIVEN_BLACKLIST = Set.new([
     "not any",
     "has not"
-  ]
+  ]).freeze
 
-  TITLE = /\s*\b(sir|count(ess)?|colonel|(gen|adm|col|maj|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+|$)/i
+  TITLE = /\s*\b(sir|count(ess)?|colonel|(gen|adm|col|maj|cmdr|lt|sgt|cpl|pvt|proff?|dr|dra\.|drª|md|ph\.?d|rev|mme|abbé|ptre|bro|esq)\.?|doct(eu|o)r|father|cantor|vicar|père|pastor|profa\.?|profª|rabbi|reverend|pere|soeur|sister|professor)(\s+|$)/i.freeze
 
-  APPELLATION = /\s*\b((mrs?|ms|fr|hr)\.?|miss|herr|frau)(\s+|$)/i
+  APPELLATION = /\s*\b((mrs?|ms|fr|hr)\.?|miss|herr|frau)(\s+|$)/i.freeze
 
-  SUFFIX = /\s*\b(JR|Jr|jr|SR|Sr|sr|ESQ|esq|[IVX]{2,})(\.|\b)/
+  SUFFIX = /\s*\b(JR|Jr|jr|SR|Sr|sr|ESQ|esq|[IVX]{2,})(\.|\b)/.freeze
 
-  PARTICLES = [
+  PARTICLES = Set.new([
     "ap",
     "da",
     "de",
@@ -425,7 +428,7 @@ module DwcAgent
     "van de",
     "van der",
     "von der"
-  ]
+  ]).freeze
 
   VOWELS = "aeiouàáâäǎæãåāèéêëěẽēėęìíîïǐĩīıįòóôöǒœøõōùúûüǔũūűů"
 

data/lib/dwc_agent/parser.rb

@@ -13,7 +13,7 @@ module DwcAgent
       subs_regex: Regexp.new(CHAR_SUBS.keys.map{|a| Regexp.escape a }.join('|').to_s),
       complex_separators_regex: COMPLEX_SEPARATORS.map{|k,v| [Regexp.new(k), v] },
       residual_terminators_regex: Regexp.new(SPLIT_BY.to_s + %r{\s*\z}.to_s)
-    }
+    }.freeze
 
     class << self
       attr_reader :defaults
@@ -35,7 +35,7 @@ module DwcAgent
     # @param names [String] the name or names to be parsed
     # @return [Array] the list of parsed names
     def parse(name)
-      return [] if name.nil? || name == ""
+      return [] if name.nil? || name.empty?
       name.gsub!(options[:strip_out_regex], ' ')
       name.gsub!(options[:tidy_remains_regex], '')
       name.gsub!(options[:subs_regex], CHAR_SUBS)

data/lib/dwc_agent/version.rb

@@ -4,7 +4,7 @@ module DwcAgent
 
     MAJOR = 3
     MINOR = 4
-    PATCH = 2
+    PATCH = 3
     BUILD = 0
 
     def self.version
@@ -12,7 +12,7 @@ module DwcAgent
     end
 
     def self.date
-      '2025-09-09'
+      '2026-02-23'
     end
 
   end

metadata

@@ -1,14 +1,13 @@
 --- !ruby/object:Gem::Specification
 name: dwc_agent
 version: !ruby/object:Gem::Version
-  version: 3.4.2.0
+  version: 3.4.3.0
 platform: ruby
 authors:
 - David P. Shorthouse
-autorequire:
 bindir: bin
 cert_chain: []
-date: 2025-09-09 00:00:00.000000000 Z
+date: 2026-02-23 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: namae
@@ -80,6 +79,20 @@ dependencies:
     - - "~>"
       - !ruby/object:Gem::Version
         version: '2'
+- !ruby/object:Gem::Dependency
+  name: benchmark-ips
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '2'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '2'
 description: Parses the typically messy content in Darwin Core terms that contain
   people names
 email: davidpshorthouse@gmail.coms
@@ -102,7 +115,6 @@ homepage: https://github.com/bionomia/dwc_agent
 licenses:
 - MIT
 metadata: {}
-post_install_message:
 rdoc_options:
 - "--encoding"
 - UTF-8
@@ -119,8 +131,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.5.4
-signing_key:
+rubygems_version: 3.6.9
 specification_version: 4
 summary: Parse Darwin Core agent terms such as recordedBy and identifiedBy
 test_files: []