filters_spam 0.1

data/lib/filters_spam.rb

@@ -0,0 +1,138 @@
+# Just call filters_spam in your model with any of the options you want.
+def filters_spam(options = {})
+  options = {
+    :message_field => :message,
+    :email_field => :email,
+    :author_field => :author,
+    :other_fields => [],
+    :extra_spam_words => []
+  }.update(options)
+
+  self.module_eval do
+    scope :ham, lambda {{:conditions => {:spam => false}, :order => 'created_at DESC'}}
+    scope :spam, lambda {{:conditions => {:spam => true}, :order => 'created_at DESC'}}
+    before_validation(:on => :create) { calculate_spam_score }
+
+    cattr_accessor :spam_words
+    self.spam_words = %w{
+      -online 4u 4-u acne adipex advicer baccarrat blackjack bllogspot booker buy byob carisoprodol
+      casino chatroom cialis coolhu credit-card-debt cwas cyclen cyclobenzaprine orgy
+      day-trading debt-consolidation discreetordering duty-free dutyfree equityloans fioricet
+      freenet free\s*shipping gambling- hair-loss homefinance holdem incest jrcreations leethal levitra macinstruct
+      mortgagequotes nemogs online-gambling ottawavalleyag ownsthis paxil penis pharmacy phentermine
+      poker poze pussy ringtones roulette shemale shoes -site slot-machine thorcarlson
+      tramadol trim-spa ultram valeofglamorganconservatives viagra vioxx xanax zolus
+    } | options[:extra_spam_words]
+  end
+
+  self.module_eval %{
+    def ham?
+      not spam?
+    end
+
+    def ham!
+      self.update_attribute(:spam, false)
+    end
+
+    def spam!
+      self.update_attribute(:spam, true)
+    end
+
+  protected
+
+    def score_for_message_links
+      link_count = self.#{options[:message_field]}.to_s.scan(/http:/).size
+      link_count > 2 ? -link_count : 2
+    end
+
+    def score_for_message_length
+      if self.#{options[:message_field]}.to_s.length > 20 and self.#{options[:message_field]}.to_s.scan(/http:/).size.zero?
+        2
+      else
+        -1
+      end
+    end
+
+    def score_for_previous_submissions
+      current_score = 0
+
+      self.class.find(:all, :conditions => {:#{options[:email_field]} => #{options[:email_field]}}).each do |i|
+        if i.spam?
+          current_score -= 1
+        else
+          current_score += 1
+        end
+      end
+
+      current_score
+    end
+
+    def score_for_spam_words
+      current_score = 0
+
+      spam_words.each do |word|
+        regex = /\#{word}/i
+        if #{options[:message_field]} =~ regex ||
+           #{options[:author_field]} =~ regex #{" || #{options[:other_fields].join(' =~ regex ')} =~ regex" if options[:other_fields].any?}
+          current_score -= 1
+        end
+      end
+
+      current_score
+    end
+
+    def score_for_suspect_url
+      current_score = 0
+
+      regex = /http:\\/\\/\\S*(\\.html|\\.info|\\?|&|free)/i
+      current_score =- (1 * #{options[:message_field]}.to_s.scan(regex).size)
+    end
+
+    def score_for_suspect_tld
+      regex = /http:\\/\\/\\S*\\.(de|pl|cn)/i
+      #{options[:message_field]}.to_s.scan(regex).size * -1
+    end
+
+    def score_for_lame_message_start
+      #{options[:message_field]}.to_s.strip =~ /^(interesting|sorry|nice|cool)/i ? -10 : 0
+    end
+
+    def score_for_author_link
+      #{options[:author_field]}.to_s.scan(/http:/).size * -2
+    end
+
+    def score_for_same_message
+      self.class.count(:conditions => {:#{options[:message_field]} => #{options[:message_field]}}) * -1
+    end
+
+    def score_for_consonant_runs
+      current_score = 0
+
+      [#{([options[:author_field], options[:message_field], options[:email_field]] | options[:other_fields]).join(', ')}].each do |field|
+        field.to_s.scan(/[bcdfghjklmnpqrstvwxz]{5,}/).each do |run|
+          current_score =- run.size - 4
+        end
+      end
+
+      current_score
+    end
+
+    def calculate_spam_score
+      score = 0
+      score += score_for_message_links
+      score += score_for_message_length
+      score += score_for_previous_submissions
+      score += score_for_spam_words
+      score += score_for_suspect_tld
+      score += score_for_lame_message_start
+      score += score_for_author_link
+      score += score_for_same_message
+      score += score_for_consonant_runs
+      self.spam = (score < 0)
+
+      logger.info("spam score was \#{score}")
+
+      true
+    end
+  }
+end

data/readme.md

@@ -0,0 +1,44 @@
+# FiltersSpam
+
+This is a small Ruby on Rails plugin that can be installed as a gem in your ``Gemfile``
+that allows models to attach to it to provide spam filtering functionality.
+
+## Rails Quickstart
+
+    # Add to Gemfile
+    gem 'filters_spam', '~> 0.1'
+
+    bundle install
+
+## Usage
+
+Once you have the plugin installed, you can use it by calling the function in your model like so:
+
+    filters_spam
+
+If you want to change the default fields that are used by ``filters_spam``
+then you can pass them in to the method as options.
+
+All of the possible options are outlined below with the default values for each:
+
+    filters_spam({
+      :message_field => :message,
+      :email_field => :email,
+      :author_field => :author,
+      :other_fields => [],
+      :extra_spam_words => %w()
+    })
+
+So, say you wanted to mark 'ruby' and 'rails' as spam words you simply pass them
+in using the ``:extra_spam_words`` option:
+
+    filters_spam({
+      :extra_spam_words => %w(ruby rails)
+    })
+
+Enjoy a life with less spam.
+
+## Credits
+
+This code was inspired by Russel Norris' [acts_as_snook plugin](http://github.com/rsl/acts_as_snook)
+and ideas presented by [Jonathan Snook](http://snook.ca/archives/other/effective_blog_comment_spam_blocker)

metadata

@@ -0,0 +1,70 @@
+--- !ruby/object:Gem::Specification 
+name: filters_spam
+version: !ruby/object:Gem::Version 
+  hash: 9
+  prerelease: false
+  segments: 
+  - 0
+  - 1
+  version: "0.1"
+platform: ruby
+authors: 
+- Philip Arndt
+- David Jones
+autorequire: 
+bindir: bin
+cert_chain: []
+
+date: 2010-08-27 00:00:00 +12:00
+default_executable: 
+dependencies: []
+
+description: This is a small Ruby on Rails plugin that can be installed as a gem in your Gemfile that allows models to attach to it to provide spam filtering functionality.
+email: info@resolvedigital.co.nz
+executables: []
+
+extensions: []
+
+extra_rdoc_files: []
+
+files: 
+- readme.md
+- lib/filters_spam.rb
+has_rdoc: true
+homepage: http://www.resolvedigital.co.nz
+licenses: []
+
+post_install_message: 
+rdoc_options: []
+
+require_paths: 
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement 
+  none: false
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      hash: 57
+      segments: 
+      - 1
+      - 8
+      - 7
+      version: 1.8.7
+required_rubygems_version: !ruby/object:Gem::Requirement 
+  none: false
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      hash: 3
+      segments: 
+      - 0
+      version: "0"
+requirements: 
+- none
+rubyforge_project: 
+rubygems_version: 1.3.7
+signing_key: 
+specification_version: 3
+summary: Attach to your model to have this filter out the spam using scoring techniques.
+test_files: []
+