voight_kampff 0.1.0

data/lib/voight_kampff.rb

@@ -0,0 +1,54 @@
+require 'voight_kampff/test'
+require 'voight_kampff/user_agents_parser'
+require 'voight_kampff/engine' if defined?(Rails)
+
+module VoightKampff
+
+  class << self
+
+    def root
+      Pathname.new File.expand_path '..', File.dirname(__FILE__)
+    end
+
+    def human?(user_agent_string)
+      test(user_agent_string).human?
+    end
+    
+    def bot?(user_agent_string)
+      test(user_agent_string).bot?
+    end
+    alias :replicant? :bot?
+
+    def browser?(user_agent_string)
+      test(user_agent_string).browser?
+    end
+
+    def checker?(user_agent_string)
+      test(user_agent_string).checker?
+    end
+    
+    def downloader?(user_agent_string)
+      test(user_agent_string).downloader?
+    end
+    
+    def proxy?(user_agent_string)
+      test(user_agent_string).proxy?
+    end
+
+    def crawler?(user_agent_string)
+      test(user_agent_string).crawler?
+    end
+
+    def spam?(user_agent_string)
+      test(user_agent_string).spam?
+    end
+
+    private
+    
+    def test(user_agent_string)
+      VoightKampff::Test.new(user_agent_string)
+    end
+
+  end
+
+end

data/lib/voight_kampff/engine.rb

@@ -0,0 +1,13 @@
+require 'voight_kampff'
+require 'rails'
+require 'active_record'
+
+module VoightKampff
+  class Engine < Rails::Engine
+  
+    rake_tasks do
+      load 'tasks/voight_kampff.rake'
+    end
+  
+  end
+end

data/lib/voight_kampff/test.rb

@@ -0,0 +1,76 @@
+module VoightKampff
+
+  class Test
+
+    attr_accessor :user_agent_string
+    attr_accessor :types
+    
+    def initialize(user_agent_string)
+      load_agents
+      @user_agent_string = user_agent_string
+      @types = Array(agent['types']).map(&:to_sym)
+      @description = agent['description']
+    end
+
+    def agent
+      @agent ||= @@agents.find do |agent|
+        if agent['string_match'] == 'regex'
+          @user_agent_string =~ Regexp.new(agent['string'], Regexp::IGNORECASE)
+        else
+          @user_agent_string.casecmp(agent['string']) == 0
+        end
+      end || {}
+    end
+
+    def has_type?(*types)
+      return nil if agent.empty?
+      types.any? do |type|
+        @types.include? type
+      end
+    end
+
+    def human?
+      has_type? :browser, :downloader, :proxy
+    end
+
+    def bot?
+      has_type? :checker, :crawler, :spam
+    end
+    alias :replicant? :bot?
+
+    def browser?
+      has_type? :browser
+    end
+
+    def checker?
+      has_type? :checker
+    end
+    
+    def downloader?
+      has_type? :downloader
+    end
+    
+    def proxy?
+      has_type? :proxy
+    end
+
+    def crawler?
+      has_type? :crawler
+    end
+
+    def spam?
+      has_type? :spam
+    end
+
+    private
+    
+    def load_agents
+      rel_path = ['config', 'user_agents.yml']
+      abs_path = Rails.root.join(*rel_path)
+      abs_path = VoightKampff.root.join(*rel_path) unless File.exists? abs_path
+      @@agents ||= YAML.load(File.open(abs_path, 'r'))
+    end
+
+  end
+
+end

data/lib/voight_kampff/user_agents_parser.rb

@@ -0,0 +1,54 @@
+module VoightKampff
+
+  class UserAgentsParser
+
+    # To generate a new user_agents.yml file run rake voight_kampff:import_user_agents
+    # This class is designed to work with http://www.user-agents.org/allagents.xml
+
+    require 'nokogiri'
+    require 'httpclient'
+
+    def initialize(url)
+      @url = url
+      @xml = HTTPClient.new.get_content(url)
+    end
+
+    def agents
+      agents = []
+      xml = Nokogiri::XML(@xml)
+      xml.xpath('//user-agent').each do |agent|
+        agents << {
+          'string' => agent.xpath('String').text,
+          'string_match' => 'exact', # exact or regex
+          'types' => agent.xpath('Type').text.split.map{|t| clarify_type(t) }.compact,
+          'description' => agent.xpath('Description').text
+        }
+      end
+      agents
+    end
+
+    def to_array
+      agents
+    end
+    alias :to_a :to_array
+
+    def to_yaml
+      agents.to_yaml
+    end
+
+    private
+    
+    def clarify_type(type)
+      case type.downcase.to_sym
+        when :b then 'browser'
+        when :c then 'checker'
+        when :d then 'downloader'
+        when :p then 'proxy'
+        when :r then 'crawler'
+        when :s then 'spam'
+      end
+    end
+
+  end
+
+end

data/lib/voight_kampff/version.rb

@@ -0,0 +1,5 @@
+module VoightKampff
+
+  VERSION = '0.1.0'
+
+end

metadata

@@ -0,0 +1,72 @@
+--- !ruby/object:Gem::Specification 
+name: voight_kampff
+version: !ruby/object:Gem::Version 
+  hash: 27
+  prerelease: 
+  segments: 
+  - 0
+  - 1
+  - 0
+  version: 0.1.0
+platform: ruby
+authors: 
+- Adam Crownoble
+autorequire: 
+bindir: bin
+cert_chain: []
+
+date: 2011-05-11 00:00:00 Z
+dependencies: []
+
+description: Voight-Kampff is a Ruby gem that detects bots, spiders, crawlers and replicants
+email: adam@obledesign.com
+executables: []
+
+extensions: []
+
+extra_rdoc_files: []
+
+files: 
+- config/user_agents.yml
+- config/initializers/extend_action_dispatch_requset.rb
+- lib/voight_kampff.rb
+- lib/voight_kampff/user_agents_parser.rb
+- lib/voight_kampff/version.rb
+- lib/voight_kampff/test.rb
+- lib/voight_kampff/engine.rb
+homepage: https://github.com/adamcrown/Voight-Kampff
+licenses: []
+
+post_install_message: 
+rdoc_options: []
+
+require_paths: 
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement 
+  none: false
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      hash: 3
+      segments: 
+      - 0
+      version: "0"
+required_rubygems_version: !ruby/object:Gem::Requirement 
+  none: false
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      hash: 3
+      segments: 
+      - 0
+      version: "0"
+requirements: []
+
+rubyforge_project: 
+rubygems_version: 1.7.2
+signing_key: 
+specification_version: 3
+summary: Voight-Kampff bot detection
+test_files: []
+
+has_rdoc: