sexmachine 0.0.3 → 0.0.4

data/Gemfile.lock

@@ -1,7 +1,7 @@
 PATH
   remote: .
   specs:
-    sexmachine (0.0.3)
+    sexmachine (0.0.4)
 
 GEM
   remote: http://rubygems.org/

data/README.rdoc

@@ -8,13 +8,22 @@ This gem uses the underlying data from the program "gender" by Jorg Michael (des
   >> d.get_gender("Sally")
   :female
   >> d.get_gender("Pauley") # should be androgynous
-  :andy 
+  :andy
+
+The result will be one of andy (androgynous), male, female, mostly_male, or mostly_female.  Any unknown names are considered andies.
 
 I18N is fully supported:
 
   >> d.get_gender("�lfr�n")
   :female
 
+Additionally, you can give preference to specific countries:
+
+  >> d.get_gender("Jamie")
+   => :female
+  >> d.get_gender("Jamie", :great_britain)
+   => :mostly_male
+
 If you have an alterative data file, you can pass that in as an optional argument to the Detector.
 
 Try to avoid creating many Detectors, as each creation means reading in the data file.

data/lib/sexmachine/detector.rb

@@ -1,6 +1,12 @@
 module SexMachine
 
   class Detector
+    COUNTRIES = [ :great_britain, :ireland, :usa, :italy, :malta, :portugal, :spain, :france, :belgium, :luxembourg, :the_netherlands, :east_frisia,
+                  :germany, :austria, :swiss, :iceland, :denmark, :norway, :sweden, :finland, :estonia, :latvia, :lithuania, :poland, :czech_republic,
+                  :slovakia, :hungary, :romania, :bulgaria, :bosniaand, :croatia, :kosovo, :macedonia, :montenegro, :serbia, :slovenia, :albania,
+                  :greece, :russia, :belarus, :moldova, :ukraine, :armenia, :azerbaijan, :georgia, :the_stans, :turkey, :arabia, :israel, :china,
+                  :india, :japan, :korea, :vietnam, :other_countries ]
+
     def initialize(fname=nil)
       fname ||= File.expand_path('../data/nam_dict.txt', __FILE__)
       parse fname
@@ -10,40 +16,68 @@ module SexMachine
       @names = {}
       open(fname, "r:iso8859-1:utf-8") { |f|
         f.each_line { |line|
-          eatNameLine line
+          eat_name_line line
         }
       }
     end
     
-    def get_gender(name)
-      @names.fetch(name, :andy)
+    def get_gender(name, country = nil)
+      if not @names.has_key?(name)
+        :andy
+      elsif country.nil?
+        most_popular_gender(name) { |country_values|
+          country_values.split("").select { |l| l.strip != "" }.length
+        }
+      elsif COUNTRIES.include?(country)
+        index = COUNTRIES.index(country)
+        most_popular_gender(name) { |country_values|
+          country_values[index].ord
+        }
+      else
+        raise "No such country: #{country}"
+      end
     end
   
     private
-    def eatNameLine(line)
+    def eat_name_line(line)
       return if line.start_with?("#") or line.start_with?("=")
       
       parts = line.split(" ").select { |p| p.strip != "" }
-      
-      if parts[0].include? "F"
-        set parts[1], :female
-      elsif parts[0].include? "M"
-        set parts[1], :male
-      else
-        set parts[1], :andy
+      country_values = line.slice(30, line.length)
+
+      case parts[0]
+        when "M" then set(parts[1], :male, country_values)
+        when "1M", "?M" then set(parts[1], :mostly_male, country_values)
+        when "F" then set(parts[1], :female, country_values)
+        when "1F", "?F" then set(parts[1], :mostly_female, country_values)
+        when "?" then set(parts[1], :andy, country_values)
+        else raise "Not sure what to do with a sex of #{parts[0]}"
       end
     end
 
-    def set(name, gender)
-      # go w/ first option, don't reset
-      return if @names.has_key? name
+    def most_popular_gender(name)
+      return :andy unless @names.has_key?(name)
       
+      max = 0
+      best = @names[name].keys.first
+      @names[name].each { |gender, country_values|
+        count = yield country_values
+        if count > max
+          max = count
+          best = gender
+        end
+      }
+      best
+    end
+
+    def set(name, gender, country_values)
       if name.include? "+"
         [ '', '-', ' ' ].each { |replacement|
-          set name.gsub("+", replacement), gender
+          set name.gsub("+", replacement), gender, country_values
         }
       else
-        @names[name] = gender
+        @names[name] ||= {}
+        @names[name][gender] = country_values
       end
     end
   end

data/lib/sexmachine/version.rb

@@ -1,3 +1,3 @@
 module SexMachine
-  VERSION = "0.0.3"
+  VERSION = "0.0.4"
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: sexmachine
 version: !ruby/object:Gem::Version
-  version: 0.0.3
+  version: 0.0.4
   prerelease: 
 platform: ruby
 authors:
@@ -9,7 +9,7 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2012-07-24 00:00:00.000000000Z
+date: 2012-07-25 00:00:00.000000000Z
 dependencies: []
 description: Get gender from first name.
 email: brian.muller@livingsocial.com