namae 0.7.0 → 0.7.1

data/features/lists.feature

@@ -61,3 +61,15 @@ Feature: Parse a list of names
       | Brian  | Kernighan |
       | Dennis | Ritchie   |
       | Donald | Knuth     |
+
+  @list @wip
+  Scenario: A list of names separated by commas
+  	Given a parser that prefers commas as separators
+    When I parse the names "G. Proctor, M. Cooper, P. Sanders & B. Malcom"
+    # Then there should be 4 names
+    Then the names should be:
+      | given | family  |
+      | G.    | Proctor |
+      | M.    | Cooper  |
+      | P.    | Sanders |
+      | B.    | Malcom  |

data/lib/namae/parser.rb

@@ -25,7 +25,7 @@ module_eval(<<'...end parser.y/module_eval...', 'parser.y', 90)
       :comma => ',',
       :separator => /\s*(\band\b|\&)\s*/i,
       :title => /\s*\b(sir|lord|(prof|dr|md|ph\.?d)\.?)(\s+|$)/i,
-      :suffix => /\s*\b(jr|sr|[ivx]+)\.?\s*/i,
+      :suffix => /\s*\b(jr|sr|[ivx]{2,})\.?\s*/i,
       :appellation => /\s*\b((mrs?|ms|fr|hr)\.?|miss|herr|frau)(\s+|$)/i
     }
   end
@@ -77,7 +77,7 @@ module_eval(<<'...end parser.y/module_eval...', 'parser.y', 90)
   end
   
   def reset
-    @commas, @words, @yydebug = 0, 0, debug?   
+    @commas, @words, @initials, @yydebug = 0, 0, 0, debug?   
     self
   end
 
@@ -93,7 +93,7 @@ module_eval(<<'...end parser.y/module_eval...', 'parser.y', 90)
   
   def consume_separator
     return next_token if seen_separator?
-    @commas, @words = 0, 0
+    @commas, @words, @initials = 0, 0, 0
     [:AND, :AND]
   end
   
@@ -104,6 +104,7 @@ module_eval(<<'...end parser.y/module_eval...', 'parser.y', 90)
 
   def consume_word(type, word)
     @words += 1
+    @initials += 1 if type == :UWORD && word =~ /^\s*[[:alpha:]]\.\s*$/
     [type, word]
   end
 
@@ -123,13 +124,13 @@ module_eval(<<'...end parser.y/module_eval...', 'parser.y', 90)
   def will_see_suffix?
     input.peek(8).to_s.strip.split(/\s+/)[0] =~ suffix
   end
-  
+    
   def will_see_initial?
     input.peek(6).to_s.strip.split(/\s+/)[0] =~ /[[:alpha:]]\./
   end
 
   def seen_full_name?
-    prefer_comma_as_separator? && @words > 1 && !will_see_initial?
+    prefer_comma_as_separator? && @words > 1 && (@initials > 0 || !will_see_initial?)
   end
 
   def next_token

data/lib/namae/parser.y

@@ -99,7 +99,7 @@ require 'strscan'
       :comma => ',',
       :separator => /\s*(\band\b|\&)\s*/i,
       :title => /\s*\b(sir|lord|(prof|dr|md|ph\.?d)\.?)(\s+|$)/i,
-      :suffix => /\s*\b(jr|sr|[ivx]+)\.?\s*/i,
+      :suffix => /\s*\b(jr|sr|[ivx]{2,})\.?\s*/i,
       :appellation => /\s*\b((mrs?|ms|fr|hr)\.?|miss|herr|frau)(\s+|$)/i
     }
   end
@@ -151,7 +151,7 @@ require 'strscan'
   end
   
   def reset
-    @commas, @words, @yydebug = 0, 0, debug?   
+    @commas, @words, @initials, @yydebug = 0, 0, 0, debug?   
     self
   end
 
@@ -167,7 +167,7 @@ require 'strscan'
   
   def consume_separator
     return next_token if seen_separator?
-    @commas, @words = 0, 0
+    @commas, @words, @initials = 0, 0, 0
     [:AND, :AND]
   end
   
@@ -178,6 +178,7 @@ require 'strscan'
 
   def consume_word(type, word)
     @words += 1
+    @initials += 1 if type == :UWORD && word =~ /^\s*[[:alpha:]]\.\s*$/
     [type, word]
   end
 
@@ -197,13 +198,13 @@ require 'strscan'
   def will_see_suffix?
     input.peek(8).to_s.strip.split(/\s+/)[0] =~ suffix
   end
-  
+    
   def will_see_initial?
     input.peek(6).to_s.strip.split(/\s+/)[0] =~ /[[:alpha:]]\./
   end
 
   def seen_full_name?
-    prefer_comma_as_separator? && @words > 1 && !will_see_initial?
+    prefer_comma_as_separator? && @words > 1 && (@initials > 0 || !will_see_initial?)
   end
 
   def next_token

data/lib/namae/version.rb

@@ -2,7 +2,7 @@ module Namae
   module Version
     MAJOR = 0
     MINOR = 7
-    PATCH = 0
+    PATCH = 1
     BUILD = nil
 
     STRING = [MAJOR, MINOR, PATCH, BUILD].compact.join('.').freeze

data/namae.gemspec

@@ -5,11 +5,11 @@
 
 Gem::Specification.new do |s|
   s.name = "namae"
-  s.version = "0.7.0"
+  s.version = "0.7.1"
 
   s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
   s.authors = ["Sylvester Keil", "Dan Collis-Puro"]
-  s.date = "2012-10-29"
+  s.date = "2012-11-08"
   s.description = " Namae (\u{540d}\u{524d}) is a parser for human names. It recognizes personal names of various cultural backgrounds and tries to split them into their component parts (e.g., given and family names, honorifics etc.). "
   s.email = ["sylvester@keil.or.at", "dan@collispuro.com"]
   s.extra_rdoc_files = [

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: namae
 version: !ruby/object:Gem::Version
-  version: 0.7.0
+  version: 0.7.1
   prerelease: 
 platform: ruby
 authors:
@@ -10,7 +10,7 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2012-10-29 00:00:00.000000000 Z
+date: 2012-11-08 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: racc
@@ -147,7 +147,7 @@ required_ruby_version: !ruby/object:Gem::Requirement
       version: '0'
       segments:
       - 0
-      hash: -219042018557465799
+      hash: -3595872270413772674
 required_rubygems_version: !ruby/object:Gem::Requirement
   none: false
   requirements: