loose_tight_dictionary 0.0.5 → 0.0.6

data/README.rdoc

@@ -1,6 +1,45 @@
 = loose_tight_dictionary
 
-Description goes here.
+Match things based on string similarity (using the Pair Distance algorithm) and regular expressions.
+
+= Quickstart
+
+    >> right_records = [ 'seamus', 'andy', 'ben' ]
+    => [...]
+    >> left_record = 'Shamus Heaney'
+    => [...]
+    >> d = LooseTightDictionary.new right_records
+    => [...]
+    >> puts d.left_to_right left_record
+    => 'seamus'
+
+Try running the included example file:
+
+    $ ruby examples/first_name_matching.rb 
+    Left side (input)
+    ====================
+    Mr. Seamus
+    Sr. Andy
+    Master BenT
+
+    Right side (output)
+    ====================
+    seamus
+    andy
+    ben
+
+    Results
+    ====================
+    Left record (input)           Right record (output)         Prefix used (if any)          Score                         
+    Mr. Seamus                    seamus                        NULL                          0.666666666666667             
+    Sr. Andy                      andy                          NULL                          0.5                           
+    Master BenT                   ben                           NULL                          0.2                           
+
+= Improving dictionaries
+
+Similarity matching will only get you so far.
+
+    TODO: regex usage
 
 == Note on Patches/Pull Requests
  

data/VERSION

@@ -1 +1 @@
-0.0.5
+0.0.6

data/examples/first_name_matching.rb

@@ -0,0 +1,23 @@
+#!/usr/bin/env ruby
+require 'rubygems'
+# require 'loose_tight_dictionary'
+require File.expand_path(File.join(File.dirname(__FILE__), '..', 'lib', 'loose_tight_dictionary.rb'))
+right_side = [ 'seamus', 'andy', 'ben' ]
+left_side = [ 'Mr. Seamus', 'Sr. Andy', 'Master BenT' ]
+
+puts "Left side (input)"
+puts "=" * 20
+puts left_side
+puts
+
+puts "Right side (output)"
+puts "=" * 20
+puts right_side
+puts
+
+puts "Results"
+puts "=" * 20
+d = LooseTightDictionary.new right_side, :tee => STDOUT, :tee_format => :fixed_width
+d.check left_side
+
+puts d.left_to_right 'Shamus Heaney'

data/lib/loose_tight_dictionary.rb

@@ -39,10 +39,11 @@ class LooseTightDictionary
   include Amatch
 
   attr_reader :right_records
-  attr_reader :logger
-  attr_reader :tee
   attr_reader :case_sensitive
-  
+
+  attr_accessor :logger
+  attr_accessor :tee
+  attr_accessor :tee_format  
   attr_accessor :positives
   attr_accessor :negatives
   attr_accessor :left_reader
@@ -59,6 +60,7 @@ class LooseTightDictionary
     @negatives = options[:negatives]
     @logger = options[:logger]
     @tee = options[:tee]
+    @tee_format = options[:tee_format] || :fixed_width
     @case_sensitive = options[:case_sensitive] || false
   end
   
@@ -106,15 +108,24 @@ class LooseTightDictionary
   end
 
   def check(left_records)
-    unless positives.present? or negatives.present?
-      logger.andand.info "You didn't define any positives or negatives, so running check doesn't do anything"
-      return
+    header = [ 'Left record (input)', 'Right record (output)', 'Prefix used (if any)', 'Score' ]
+    case tee_format
+    when :csv
+      tee.andand.puts header.flatten.to_csv
+    when :fixed_width
+      tee.andand.puts header.map { |i| i.to_s.ljust(30) }.join
     end
+
     left_records.each do |left_record|
       begin
         right_record = left_to_right left_record
       ensure
-        tee.andand.puts [ read_left(left_record), read_right($ltd_0), $ltd_1 ].flatten.to_csv
+        case tee_format
+        when :csv
+          tee.andand.puts $ltd_1.flatten.to_csv
+        when :fixed_width
+          tee.andand.puts $ltd_1.map { |i| i.to_s.ljust(30) }.join if $ltd_1
+        end
       end
     end
   end
@@ -292,12 +303,24 @@ class LooseTightDictionary
   
   def read_left(left_record)
     return if left_record.nil?
-    left_reader ? left_reader.call(left_record) : left_record[0]
+    if left_reader
+      left_reader.call(left_record)
+    elsif left_record.is_a?(String)
+      left_record
+    else
+      left_record[0]
+    end
   end
   
   def read_right(right_record)
     return if right_record.nil?
-    right_reader ? right_reader.call(right_record) : right_record[0]
+    if right_reader
+      right_reader.call(right_record)
+    elsif right_record.is_a?(String)
+      right_record
+    else
+      right_record[0]
+    end
   end
   
   # Thanks William James!

data/test/test_loose_tight_dictionary.rb

@@ -68,6 +68,28 @@ class TestLooseTightDictionary < Test::Unit::TestCase
   end
   
   if ENV['OLD'] == 'true' or ENV['ALL'] == 'true'
+    # the example from the readme, considerably uglier here
+    should "check a simple table" do
+      @right = [ 'seamus', 'andy', 'ben' ]
+      @positives = [ [ 'seamus', 'Mr. Seamus Abshere' ] ]
+      left = [ 'Mr. Seamus Abshere', 'Sr. Andy Rossmeissl', 'Master BenT' ]
+    
+      assert_nothing_raised do
+        ltd.check left
+      end
+    end
+    
+    should "treat a String as a full record if passed through" do
+      dash = 'DHC8-400'
+      b747 = 'B747200/300'
+      dc9 = 'DC-9-10'
+      right_records = [ dash, b747, dc9 ]
+      simple_ltd = LooseTightDictionary.new right_records, :logger => $logger, :tee => $tee
+      assert_equal dash, simple_ltd.left_to_right('DeHavilland Dash-8 DHC-400')
+      assert_equal b747, simple_ltd.left_to_right('Boeing 747-300')
+      assert_equal dc9, simple_ltd.left_to_right('McDonnell Douglas MD81/DC-9')
+    end
+    
     should "call it a mismatch if you hit a blank positive" do
       @positives.push [@a_left[0], '']
       assert_raises(LooseTightDictionary::Mismatch) do

metadata

@@ -5,8 +5,8 @@ version: !ruby/object:Gem::Version
   segments: 
   - 0
   - 0
-  - 5
-  version: 0.0.5
+  - 6
+  version: 0.0.6
 platform: ruby
 authors: 
 - Seamus Abshere
@@ -14,7 +14,7 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2010-05-03 00:00:00 -04:00
+date: 2010-05-13 00:00:00 -04:00
 default_executable: 
 dependencies: 
 - !ruby/object:Gem::Dependency 
@@ -115,6 +115,7 @@ files:
 - README.rdoc
 - Rakefile
 - VERSION
+- examples/first_name_matching.rb
 - examples/icao-bts.rb
 - examples/icao-bts.xls
 - lib/loose_tight_dictionary.rb
@@ -153,4 +154,5 @@ summary: Allows iterative development of dictionaries for big data sets.
 test_files: 
 - test/helper.rb
 - test/test_loose_tight_dictionary.rb
+- examples/first_name_matching.rb
 - examples/icao-bts.rb