namor 0.0.1 → 0.0.2

data/.gitignore

@@ -15,3 +15,4 @@ spec/reports
 test/tmp
 test/version_tmp
 tmp
+.DS_Store

data/lib/namor/namor.rb

@@ -0,0 +1,77 @@
+class Namor::Namor
+  def initialize(opts = {})
+    config(opts)
+  end
+
+  def config(opts)
+    @config = opts
+  end
+
+  def extract(name)
+    return [] if name.nil?
+
+    suppression_list = @config[:suppress] || []
+    suppression_re = suppression_list.join('|')
+
+    detitled_name = name.upcase.gsub(/\b(#{suppression_re})\b/i, '').gsub(/\b(MD|JR|SR|I+|IV)\b/i, '').gsub(/\([^\(]*\)/, '').gsub(/[_.'-]/, '').gsub(/,\s*$/, '').gsub(/ +/, ' ').strip
+
+    if detitled_name =~ /,/
+      # "last, first[ middle]"
+      lastname, firstname = detitled_name.split(/\s*,\s*/)
+      lastname.gsub!(/ /, '')
+      middlename = nil
+      if firstname && firstname =~ / /
+        pieces = firstname.split(/ +/)
+        firstname = pieces.shift
+        middlename = pieces.join if pieces.any?
+      end
+    else
+      # "first [middle ]last"
+      pieces = detitled_name.split(' ')
+      firstname = pieces.shift
+      middlename = nil
+      if pieces.count > 1 && pieces.first.length == 1
+        # assume this is a middle initial
+        middlename = pieces.shift
+      end
+
+      lastname = pieces.join
+    end
+
+    firstname = nil if firstname.empty?
+    middlename = nil if middlename && middlename.empty?
+    lastname = nil if lastname.empty?
+
+    fm = [firstname, middlename].compact.join(' ')
+    fullname = [lastname, fm].compact.join(',')
+
+    [firstname, middlename, lastname, fullname]
+  end
+
+  def extract_with_cluster(name)
+    ary = extract(name)
+    return [] if ary.empty?
+    ary << ary.last.gsub(/\W/, '_')
+  end
+
+
+  def components(*args)
+    suppression_list = @config[:suppress] ? @config[:suppress].map(&:upcase) : []
+
+    names = args
+    bits = []
+    names.compact.each do |name|
+      name = name.dup
+      name.gsub!(/\([^\(]*\)/, '')
+      name.gsub!(/\[[^\[]*\]/, '')
+      name.gsub!(/[\(\)\[\]\']/, '')
+      name.gsub!(/[,._-]/, ' ')
+      bits += name.split(/\s+/).map(&:upcase)
+    end
+
+    suppress_re = %w{MD JR SR I+ IV}.join('|')
+    bits.delete_if {|bit| suppression_list.include?(bit) || bit =~ /^(#{suppress_re})$/}
+    bits.delete_if(&:empty?)
+    bits.uniq.sort
+  end
+end

data/lib/namor/version.rb

@@ -1,3 +1,3 @@
 module Namor
-  VERSION = "0.0.1"
+  VERSION = "0.0.2"
 end

data/lib/namor.rb

@@ -1,50 +1,5 @@
-require "namor/version"
+require_relative "namor/version"
+require_relative "namor/namor"
 
 module Namor
-  def self.extract(name, args = {})
-    return [] if name.nil?
-
-    suppression_list = args[:suppress] || []
-    suppression_re = suppression_list.join('|')
-
-    detitled_name = name.upcase.gsub(/\b(#{suppression_re})\b/i, '').gsub(/\b(MD|JR|SR|I+|IV)\b/i, '').gsub(/\([^\(]*\)/, '').gsub(/[_.'-]/, '').gsub(/,\s*$/, '').gsub(/ +/, ' ').strip
-
-    if detitled_name =~ /,/
-      # "last, first[ middle]"
-      lastname, firstname = detitled_name.split(/\s*,\s*/)
-      lastname.gsub!(/ /, '')
-      middlename = nil
-      if firstname && firstname =~ / /
-        pieces = firstname.split(/ +/)
-        firstname = pieces.shift
-        middlename = pieces.join if pieces.any?
-      end
-    else
-      # "first [middle ]last"
-      pieces = detitled_name.split(' ')
-      firstname = pieces.shift
-      middlename = nil
-      if pieces.count > 1 && pieces.first.length == 1
-        # assume this is a middle initial
-        middlename = pieces.shift
-      end
-
-      lastname = pieces.join
-    end
-
-    firstname = nil if firstname.empty?
-    middlename = nil if middlename && middlename.empty?
-    lastname = nil if lastname.empty?
-
-    fm = [firstname, middlename].compact.join(' ')
-    fullname = [lastname, fm].compact.join(',')
-
-    [firstname, middlename, lastname, fullname]
-  end
-
-  def self.extract_with_cluster(name, args = {})
-    ary = extract(name, args)
-    return [] if ary.empty?
-    ary << ary.last.gsub(/\W/, '_')
-  end
 end

data/spec/lib/namor_spec.rb

@@ -3,61 +3,92 @@
 require "spec_helper"
 
 describe "name extract" do
+  before(:all) do
+    @namor = Namor::Namor.new
+    @namor.config(:suppress => ['MD', 'DDS'])
+  end
+
   it "should handle 2-part names without commas" do
-    Namor::extract("john smith").should == ['JOHN', nil, 'SMITH', 'SMITH,JOHN']
+    @namor.extract("john smith").should == ['JOHN', nil, 'SMITH', 'SMITH,JOHN']
   end
 
   it "should handle 2-part names with commas" do
-    Namor::extract("SMITH, JOHN").should == ['JOHN', nil, 'SMITH', 'SMITH,JOHN']
+    @namor.extract("SMITH, JOHN").should == ['JOHN', nil, 'SMITH', 'SMITH,JOHN']
   end
 
   it "should handle 2-part names with commas and middle initials" do
-    Namor::extract("SMITH, JOHN R").should == ['JOHN', 'R', 'SMITH', 'SMITH,JOHN R']
+    @namor.extract("SMITH, JOHN R").should == ['JOHN', 'R', 'SMITH', 'SMITH,JOHN R']
   end
 
   it "should handle 2-part names with commas and middle initials" do
-    Namor::extract("SMITH, JOHN R").should == ['JOHN', 'R', 'SMITH', 'SMITH,JOHN R']
+    @namor.extract("SMITH, JOHN R").should == ['JOHN', 'R', 'SMITH', 'SMITH,JOHN R']
   end
 
   it "should strip elements within parentheses" do
-    Namor::extract("SMITH, JOHN (Jacko) R").should == ['JOHN', 'R', 'SMITH', 'SMITH,JOHN R']
+    @namor.extract("SMITH, JOHN (Jacko) R").should == ['JOHN', 'R', 'SMITH', 'SMITH,JOHN R']
   end
 
   it "should drop periods" do
-    Namor::extract("John R. Smith").should == ['JOHN', 'R', 'SMITH', 'SMITH,JOHN R']
+    @namor.extract("John R. Smith").should == ['JOHN', 'R', 'SMITH', 'SMITH,JOHN R']
   end
 
   it "should drop spaces in last name (only when input has a comma)" do
-    Namor::extract("Smith Jones, Mary").should == ['MARY', nil, 'SMITHJONES', 'SMITHJONES,MARY']
+    @namor.extract("Smith Jones, Mary").should == ['MARY', nil, 'SMITHJONES', 'SMITHJONES,MARY']
   end
 
   it "should drop dashes & apostrophes" do
-    Namor::extract("Mary Smith-Jones").should == ['MARY', nil, 'SMITHJONES', 'SMITHJONES,MARY']
-    Namor::extract("Mary S. O'Keefe").should == ['MARY', 'S', 'OKEEFE', 'OKEEFE,MARY S']
-    Namor::extract("Jean-Michel Claude").should == ['JEANMICHEL', nil, 'CLAUDE', 'CLAUDE,JEANMICHEL']
+    @namor.extract("Mary Smith-Jones").should == ['MARY', nil, 'SMITHJONES', 'SMITHJONES,MARY']
+    @namor.extract("Mary S. O'Keefe").should == ['MARY', 'S', 'OKEEFE', 'OKEEFE,MARY S']
+    @namor.extract("Jean-Michel Claude").should == ['JEANMICHEL', nil, 'CLAUDE', 'CLAUDE,JEANMICHEL']
   end
 
   it "should concatenate extract name pieces" do
-    Namor::extract("rajesh kumar vishnu garuda").should == ['RAJESH', nil, 'KUMARVISHNUGARUDA', 'KUMARVISHNUGARUDA,RAJESH']
-    Namor::extract("Kumar, Rajesh Vishnu Garuda").should == ['RAJESH', 'VISHNUGARUDA', 'KUMAR', 'KUMAR,RAJESH VISHNUGARUDA']
+    @namor.extract("rajesh kumar vishnu garuda").should == ['RAJESH', nil, 'KUMARVISHNUGARUDA', 'KUMARVISHNUGARUDA,RAJESH']
+    @namor.extract("Kumar, Rajesh Vishnu Garuda").should == ['RAJESH', 'VISHNUGARUDA', 'KUMAR', 'KUMAR,RAJESH VISHNUGARUDA']
   end
 
   it "should excise suffixes like 'Jr.' from lastnames" do
-    Namor::extract("Smith Jr, Edward M").should == ['EDWARD', 'M', 'SMITH', 'SMITH,EDWARD M']
+    @namor.extract("Smith Jr, Edward M").should == ['EDWARD', 'M', 'SMITH', 'SMITH,EDWARD M']
   end
 
   it "should excise terms from optional suppression list" do
-    Namor::extract("Smith Jr, Edward M MD DDS", :suppress => ['MD', 'DDS']).should == ['EDWARD', 'M', 'SMITH', 'SMITH,EDWARD M']
-    Namor::extract("Smith Jr, Edward III MD PHD", :suppress => ['MD', 'DDS']).should == ['EDWARD', 'PHD', 'SMITH', 'SMITH,EDWARD PHD']
+    @namor.extract("Smith Jr, Edward M MD DDS").should == ['EDWARD', 'M', 'SMITH', 'SMITH,EDWARD M']
+    @namor.extract("Smith Jr, Edward III MD PHD").should == ['EDWARD', 'PHD', 'SMITH', 'SMITH,EDWARD PHD']
   end
 
   it "should handle pathological cases" do
-    Namor::extract(", Mary Smith").should == ['MARY', 'SMITH', nil, 'MARY SMITH']
+    @namor.extract(", Mary Smith").should == ['MARY', 'SMITH', nil, 'MARY SMITH']
   end
 end
 
 describe "with cluster coding" do
+  before(:all) do
+    @namor = Namor::Namor.new
+    @namor.config(:suppress => ['MD', 'DDS'])
+  end
+
   it "should generate cluster labels" do
-    Namor::extract_with_cluster("Smith Jr, Edward III MD PHD", :suppress => ['MD', 'DDS']).last.should == 'SMITH_EDWARD_PHD'
+    @namor.extract_with_cluster("Smith Jr, Edward III MD PHD").last.should == 'SMITH_EDWARD_PHD'
+  end
+end
+
+describe "name componentization" do
+  before(:all) do
+    @namor = Namor::Namor.new
+    @namor.config(:suppress => ['esq'])
+  end
+
+  it "should include initials" do
+    @namor.components("john q. smith").should == ['JOHN', 'Q', 'SMITH']
+  end
+
+  it "should excise common suffixes" do
+    @namor.components("john smith III").should == ['JOHN', 'SMITH']
+    @namor.components("john smith jr").should == ['JOHN', 'SMITH']
+  end
+
+  it "should excise from suppression list" do
+    @namor.components("john smith esk.").should == ['ESK', 'JOHN', 'SMITH']
+    @namor.components("john smith esq.").should == ['JOHN', 'SMITH']
   end
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: namor
 version: !ruby/object:Gem::Version
-  version: 0.0.1
+  version: 0.0.2
   prerelease: 
 platform: ruby
 authors:
@@ -9,11 +9,11 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2012-08-08 00:00:00.000000000 Z
+date: 2012-08-23 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rake
-  requirement: &2152651100 !ruby/object:Gem::Requirement
+  requirement: !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ~>
@@ -21,10 +21,15 @@ dependencies:
         version: 0.9.2
   type: :development
   prerelease: false
-  version_requirements: *2152651100
+  version_requirements: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: 0.9.2
 - !ruby/object:Gem::Dependency
   name: rspec
-  requirement: &2152661440 !ruby/object:Gem::Requirement
+  requirement: !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ~>
@@ -32,10 +37,15 @@ dependencies:
         version: 2.9.0
   type: :development
   prerelease: false
-  version_requirements: *2152661440
+  version_requirements: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: 2.9.0
 - !ruby/object:Gem::Dependency
   name: guard-rspec
-  requirement: &2152659940 !ruby/object:Gem::Requirement
+  requirement: !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ~>
@@ -43,10 +53,15 @@ dependencies:
         version: 0.7.0
   type: :development
   prerelease: false
-  version_requirements: *2152659940
+  version_requirements: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: 0.7.0
 - !ruby/object:Gem::Dependency
   name: ruby_gntp
-  requirement: &2152658740 !ruby/object:Gem::Requirement
+  requirement: !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ~>
@@ -54,7 +69,12 @@ dependencies:
         version: 0.3.4
   type: :development
   prerelease: false
-  version_requirements: *2152658740
+  version_requirements: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: 0.3.4
 description: Munging English names
 email:
 - jmay@pobox.com
@@ -68,6 +88,7 @@ files:
 - README.md
 - Rakefile
 - lib/namor.rb
+- lib/namor/namor.rb
 - lib/namor/version.rb
 - namor.gemspec
 - spec/lib/namor_spec.rb
@@ -84,19 +105,24 @@ required_ruby_version: !ruby/object:Gem::Requirement
   - - ! '>='
     - !ruby/object:Gem::Version
       version: '0'
+      segments:
+      - 0
+      hash: 3634688006053498514
 required_rubygems_version: !ruby/object:Gem::Requirement
   none: false
   requirements:
   - - ! '>='
     - !ruby/object:Gem::Version
       version: '0'
+      segments:
+      - 0
+      hash: 3634688006053498514
 requirements: []
 rubyforge_project: 
-rubygems_version: 1.8.10
+rubygems_version: 1.8.24
 signing_key: 
 specification_version: 3
 summary: Parse & extract pieces of names
 test_files:
 - spec/lib/namor_spec.rb
 - spec/spec_helper.rb
-has_rdoc: