Parses a postal address into its components. (limited) i18n support and other features. inspired by this forrst post: http://forr.st/~zDa

PostalAddress.rb

#
# Most of the regexes here are from http://twitter.com/schuyler.
# They were posted on Forrst (http://forr.st/~zDa) as a public post,
# so I reworked them into a full class.
# 
# Added:
# => (limited) international support -- should handle US/Canada and most of western europe
# => added :receipient, :postcode, and :country fields
# => state/province detection built-in

require 'yaml'

class PostalAddress
  
  attr_accessor :addr
  
  Match = {
      :number     => /^(\d+\W|[a-z]+)?(\d+)([a-z]?)\b/io,
      :unit       => /\W?(?:Ste|Suite|#|Unit|Apt|Apartment)\s?[0-9]+/,
      :street     => 
      {
        :regex => /(?:\b(?:\d+\w*|[a-z'-]+)\s*)+/io, 
        :block => Proc.new {|v| v.lines.select {|l| l.match(/\A[0-9]/) }.first.strip } 
      },
      :city       => 
      {
        :regex => /(?:\b(?:\d+\w*|[a-z'-]+)\s*)+/io,
        :block => Proc.new {|v| v.lines.last }
      },
      :state      => 
      {
        :regex => /\W(A[BLKSZRAEP]|BC|C[AOT]|D[EC]|F[LM]|G[ANU]|HI|I[ADLN]|K[SY]|LA|M[ABDEHINOPST]|N[BCDEHJLMSUVY]|O[HKNR]|P[AERW]|QC|RI|S[CDK]|T[NX]|UT|V[AIT]|W[AIVY]|YT)\s/i,
        :subexpression => 1
      },
      :zip   => /(\d{5})(?:-\d{4})?\s*$/o,
      :postcode   => /(?:(?:[A-Z0-9]{3,}) (?:[A-Z0-9]{3,})\s*$|(\d{5})(?:-\d{4})?\s*$)/o,
      :at         => /\s(at|@|and|&)\s/io,
      :po_box     => /\b[P|p]*(OST|ost)*\.*\s*[O|o|0]*(ffice|FFICE)*\.*\s*[B|b][O|o|0][X|x]\b/,
      :recipient  => /\A[a-zA-Z'\s]+$/iox,
      :country    => /^[A-Z\s]+\z/io
  }
  
  def initialize(addr)
    @addr = addr
  end
  
  def country
    return "Canada" if /AB|BC|MB|N[BLTSU]|ON|PE|QC|SK|YT/.match(self.state)
    return "USA" if Match[:state][:regex].match(@addr)
    return "United Kingdom" if self.postcode
    self.class.send(:country, @addr)
  end
  
  def method_missing(name)
    self.class.method_missing(name, @addr)
  end
  
  def self.method_missing(name, addr)
    m = Match[name]
    if m.is_a?(Hash) 
      matched = m[:regex].match(addr)
      if !matched.nil?
        if m[:subexpression]
          result = matched[m[:subexpression]].to_s.strip
        end
        if m[:block]
          result ||= matched.to_s.strip
          result = m[:block].call(result)
        end
        return result if !result.nil?
      end
      m = m[:regex] # Fall through
    end
    result = m.match(addr).to_s.strip
    result = result.to_i if result.is_numeric?
    result
  end
  
  def self.keys
    Match.keys
  end
  
  def to_hash
    h = {}
    self.class.keys.each do |part|
      h[part.to_s] = self.class.send(part, @addr)
    end
    h
  end
  
  def to_yaml
    to_hash.to_yaml
  end
  
end

class String
  
  def is_numeric? # "222" or "4.5" counts
    !(self.match(/\A[+-]?\d+?(\.\d+)?\Z/) == nil)
  end
  
  def lines
    self.split("\n").to_a
  end
  
end

puts "US Address"

addr = "Colin Young
222 N Ashland Suite 300
Chicago, IL 60657"

puts PostalAddress.new(addr).to_yaml
puts "Detected Country: #{PostalAddress.new(addr).country}"

puts "\n\nCanadian Address"

addr = "230 Riot Boulevard
Vancouver, BC V5Y 1K6"

puts PostalAddress.new(addr).to_yaml
puts "Detected Country: #{PostalAddress.new(addr).country}"

puts "\n\nUK Address"

addr = "Colin Young
23 Herefordshire Ln
LONDON
SW1 P01"

puts PostalAddress.new(addr).to_yaml
puts "Detected Country: #{PostalAddress.new(addr).country}"

Awesome

Awesome