duns-lookup 0.0.3

data/duns-lookup.gemspec

@@ -0,0 +1,71 @@
+# Generated by jeweler
+# DO NOT EDIT THIS FILE
+# Instead, edit Jeweler::Tasks in Rakefile, and run `rake gemspec`
+# -*- encoding: utf-8 -*-
+
+Gem::Specification.new do |s|
+  s.name = %q{duns-lookup}
+  s.version = "0.0.3"
+
+  s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
+  s.authors = ["Kali Donovan"]
+  s.date = %q{2009-10-14}
+  s.description = %q{Provides a small wrapper around the Dun & Bradstreet website to retrieve business information from DUNS numbers.}
+  s.email = %q{kali.donovan@gmail.com}
+  s.extra_rdoc_files = [
+    "LICENSE",
+     "README.rdoc"
+  ]
+  s.files = [
+    ".document",
+     ".gitignore",
+     "LICENSE",
+     "README.rdoc",
+     "Rakefile",
+     "VERSION",
+     "doc/classes/Duns.html",
+     "doc/classes/Duns.src/M000001.html",
+     "doc/classes/Duns.src/M000002.html",
+     "doc/classes/Duns.src/M000003.html",
+     "doc/classes/Duns.src/M000004.html",
+     "doc/classes/Duns.src/M000005.html",
+     "doc/classes/Duns.src/M000006.html",
+     "doc/classes/DunsError.html",
+     "doc/created.rid",
+     "doc/files/lib/duns-lookup_rb.html",
+     "doc/fr_class_index.html",
+     "doc/fr_file_index.html",
+     "doc/fr_method_index.html",
+     "doc/index.html",
+     "doc/rdoc-style.css",
+     "duns-lookup.gemspec",
+     "lib/duns-lookup.rb",
+     "spec/duns-lookup_spec.rb",
+     "spec/spec_helper.rb"
+  ]
+  s.homepage = %q{http://github.com/kdonovan/duns-lookup}
+  s.rdoc_options = ["--charset=UTF-8"]
+  s.require_paths = ["lib"]
+  s.rubygems_version = %q{1.3.5}
+  s.summary = %q{Wrapper for looking up Dun & Bradstreet numbers (DUNS)}
+  s.test_files = [
+    "spec/duns-lookup_spec.rb",
+     "spec/spec_helper.rb"
+  ]
+
+  if s.respond_to? :specification_version then
+    current_version = Gem::Specification::CURRENT_SPECIFICATION_VERSION
+    s.specification_version = 3
+
+    if Gem::Version.new(Gem::RubyGemsVersion) >= Gem::Version.new('1.2.0') then
+      s.add_development_dependency(%q<rspec>, [">= 0"])
+      s.add_runtime_dependency(%q<mechanize>, [">= 0"])
+    else
+      s.add_dependency(%q<rspec>, [">= 0"])
+      s.add_dependency(%q<mechanize>, [">= 0"])
+    end
+  else
+    s.add_dependency(%q<rspec>, [">= 0"])
+    s.add_dependency(%q<mechanize>, [">= 0"])
+  end
+end

data/lib/duns-lookup.rb

@@ -0,0 +1,99 @@
+class DunsError < Exception
+end
+
+# = Synopsis
+# The Duns library provides a small wrapper around the Dun & Bradstreet
+# website's advanced search functionality.  Currently it only implements
+# searching for company by DUNS number.
+#
+# == Example
+#  require 'rubygems'
+#  require 'kdonovan-duns-lookup'
+#
+#  Duns.lookup_duns( *invalid_number* )
+#     # => nil
+#  Duns.lookup_duns( *some_valid_number* )
+#     # => {:name => *a_name*, :address => *an_address*}
+
+class Duns
+  require 'rubygems'
+  require 'open-uri'
+  require 'mechanize'
+  
+  # D&B homepage URL
+  @@dnb_homepage = 'http://smallbusiness.dnb.com/'
+  
+  # The URI for D&B advanced search
+  @@dnb_advanced_search = 'https://smallbusiness.dnb.com/ePlatform/servlet/AdvancedCompanySearch?storeId=10001&catalogId=70001?storeId=10001&catalogId=70001'
+
+  # Our Mechanize agent
+  @@agent = WWW::Mechanize.new
+  @@agent.user_agent_alias = 'Windows IE 7'
+
+  ####
+  # Look up a given DUNS number in the D&B database.
+  # If the number is found, returns a hash with :name and :address keys.
+  # Otherwise, returns nil.
+  #  
+  def self.lookup_duns(number)
+    form = @@agent.get( @@dnb_advanced_search ).form('DunsSearchForm')
+    form.dunsNumber = enforce_duns_formatting(number)
+    page = @@agent.submit(form)
+    
+    extract_search_results(page)
+  end
+
+
+  ####
+  # Updates the internal URL used as the base for advanced searches.
+  # 
+  # The D&B website uses lots of extraneous (to us) URL params, and I have no idea what they all mean.
+  # If the base URL stops working, this method will try to set a new one by visiting the main page and
+  # finding & clicking an Advanced Search link.
+  # 
+  # This is mostly a precaution, but if searches stop working a good first bet would be to try running this
+  # method.  If search is still broken, something actually changed in the D&B HTML and we'll need to retool
+  # the gem.
+  #
+  def self.update_advanced_search_url
+    page = @@agent.get( @@dnb_homepage )
+    advanced_link = page.links.find{|l| l.text.match(/search/i) && l.uri.to_s.match(/AdvancedCompanySearch/)}
+    raise(DunsError, "Unable to find an advanced search link at: #{@@dnb_homepage}") if advanced_link.nil?
+    @@dnb_advanced_search = advanced_link.uri.to_s
+  end
+  
+  
+  protected
+  
+  # Given a search page, extract the results
+  def self.extract_search_results(page)
+    return nil if page.search('div.text-red').size > 0
+    
+    # Given a DUNS number search, we only expect one result.  For other search types (when/if implemented), loop over all TRs returned in this table
+    company_name    = extract_name( page.search("//table[@id='SearchResultsTable']//tr[1]/td[2]") )
+    company_address = extract_address( page.search("//table[@id='SearchResultsTable']//tr[1]/td[3]") )
+
+    {:name => company_name, :address => company_address}
+  end
+  
+  # Retrieve an address from the proper D&B HTML for the td element
+  def self.extract_address(td)
+    td.text.gsub(/\s{2,}/, ' ')
+  end
+  
+  # Retrieve a name from the proper D&B HTML for the td element
+  def self.extract_name(td)
+    raw_text = td.text.gsub(/\s{2,}/, ' ')
+    with_js = raw_text.match(/var companyName=escape(.+?)var companyAddr/)[1] # First get the surrounding js, to ensure we don't get confused by e.g. a ) in the company name
+
+    # Now we have e.g. >> ('NORTH TEXAS CIRCUIT BOARD CO., INC.'); << and we need to strip off the extraneous stuff
+    with_js[2..-5]
+  end
+  
+  # Strip out non-numeric characters, and raise error if remaining number is still invalid
+  def self.enforce_duns_formatting(orig_number)
+    number = orig_number.to_s.gsub(/\D/, '')
+    raise(DunsError, "Received invalid DUNS number (must be 9 digits): #{orig_number}") unless number.length == 9
+    return number
+  end
+end

data/spec/duns-lookup_spec.rb

@@ -0,0 +1,35 @@
+require File.expand_path(File.dirname(__FILE__) + '/spec_helper')
+
+describe "DunsLookup" do
+  context "Searching by DUNS number" do
+    it "should raise an exception when invalid DUNS numbers are supplied" do
+      lambda { Duns.lookup_duns('123') }.should raise_error(DunsError)
+      lambda { Duns.lookup_duns('abcdefghi') }.should raise_error(DunsError)
+      lambda { Duns.lookup_duns('123abc---') }.should raise_error(DunsError)
+      lambda { Duns.lookup_duns('') }.should raise_error(DunsError)
+    end
+
+    it "should not raise an exception when a valid DUNS number is supplied" do
+      lambda { Duns.lookup_duns('123456789') }.should_not raise_error(DunsError)
+      lambda { Duns.lookup_duns('123 456 789') }.should_not raise_error(DunsError)
+      lambda { Duns.lookup_duns('123-456-789') }.should_not raise_error(DunsError)
+    end
+    
+    it "Returns no results where there shouldn't be any" do
+      Duns.lookup_duns( fake_duns[:number] ).should be_nil
+    end
+  
+    it "Should properly extract company name and address when a valid result is found" do
+      unless real_duns[:number] && real_duns[:name] && real_duns[:address]
+        raise "Must provide information from an existing company to validates results are correct. Edit #real_duns in spec_helper.rb"
+      end
+
+      results = Duns.lookup_duns( real_duns[:number] )
+      results.should_not be_nil
+      results.should be_a_kind_of Hash
+      
+      results[:name].should == real_duns[:name]
+      results[:address].should == real_duns[:address]
+    end
+  end
+end

data/spec/spec_helper.rb

@@ -0,0 +1,25 @@
+$LOAD_PATH.unshift(File.dirname(__FILE__))
+$LOAD_PATH.unshift(File.join(File.dirname(__FILE__), '..', 'lib'))
+require 'duns-lookup'
+require 'spec'
+require 'spec/autorun'
+
+Spec::Runner.configure do |config|
+  
+end
+
+# TODO -- to run the specs, input valid information from a real company here
+# Try looking up your business by hand first to see exactly the format used by the D&B website
+# (for instance punctuation matters, as in the abbreviations of CO and INC below)
+def real_duns
+  {
+    :number   => nil,  # Enter the company's DUNS number
+    :name     => nil,  # Enter the official company name, ALL UPPER CASE: FOO WIDGET CO., INC.
+    :address  => nil   # Enter the officially registered address, ALL UPPER CASE: 123 W MAIN ST, SAN FRANCISCO, CA
+  }
+end
+
+# A valid, but non-existant DUNS number
+def fake_duns
+  {:number => '123456789'}
+end

metadata

@@ -0,0 +1,100 @@
+--- !ruby/object:Gem::Specification 
+name: duns-lookup
+version: !ruby/object:Gem::Version 
+  version: 0.0.3
+platform: ruby
+authors: 
+- Kali Donovan
+autorequire: 
+bindir: bin
+cert_chain: []
+
+date: 2009-10-14 00:00:00 -07:00
+default_executable: 
+dependencies: 
+- !ruby/object:Gem::Dependency 
+  name: rspec
+  type: :development
+  version_requirement: 
+  version_requirements: !ruby/object:Gem::Requirement 
+    requirements: 
+    - - ">="
+      - !ruby/object:Gem::Version 
+        version: "0"
+    version: 
+- !ruby/object:Gem::Dependency 
+  name: mechanize
+  type: :runtime
+  version_requirement: 
+  version_requirements: !ruby/object:Gem::Requirement 
+    requirements: 
+    - - ">="
+      - !ruby/object:Gem::Version 
+        version: "0"
+    version: 
+description: Provides a small wrapper around the Dun & Bradstreet website to retrieve business information from DUNS numbers.
+email: kali.donovan@gmail.com
+executables: []
+
+extensions: []
+
+extra_rdoc_files: 
+- LICENSE
+- README.rdoc
+files: 
+- .document
+- .gitignore
+- LICENSE
+- README.rdoc
+- Rakefile
+- VERSION
+- doc/classes/Duns.html
+- doc/classes/Duns.src/M000001.html
+- doc/classes/Duns.src/M000002.html
+- doc/classes/Duns.src/M000003.html
+- doc/classes/Duns.src/M000004.html
+- doc/classes/Duns.src/M000005.html
+- doc/classes/Duns.src/M000006.html
+- doc/classes/DunsError.html
+- doc/created.rid
+- doc/files/lib/duns-lookup_rb.html
+- doc/fr_class_index.html
+- doc/fr_file_index.html
+- doc/fr_method_index.html
+- doc/index.html
+- doc/rdoc-style.css
+- duns-lookup.gemspec
+- lib/duns-lookup.rb
+- spec/duns-lookup_spec.rb
+- spec/spec_helper.rb
+has_rdoc: true
+homepage: http://github.com/kdonovan/duns-lookup
+licenses: []
+
+post_install_message: 
+rdoc_options: 
+- --charset=UTF-8
+require_paths: 
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement 
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      version: "0"
+  version: 
+required_rubygems_version: !ruby/object:Gem::Requirement 
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      version: "0"
+  version: 
+requirements: []
+
+rubyforge_project: 
+rubygems_version: 1.3.5
+signing_key: 
+specification_version: 3
+summary: Wrapper for looking up Dun & Bradstreet numbers (DUNS)
+test_files: 
+- spec/duns-lookup_spec.rb
+- spec/spec_helper.rb