spamham 0.0.2

data/.gitignore

@@ -0,0 +1,5 @@
+pkg/*
+*.gem
+.bundle
+*~
+tmp/*

data/Gemfile

@@ -0,0 +1,4 @@
+source "http://rubygems.org"
+
+# Specify your gem's dependencies in spamham.gemspec
+gemspec

data/README.rdoc

@@ -0,0 +1,3 @@
+= SpamHam
+
+SpamHam provides a simple method for and string, .spam? this simply returns true or flase is SpamHam belives the message to be spam

data/Rakefile

@@ -0,0 +1,2 @@
+require 'bundler'
+Bundler::GemHelper.install_tasks

data/init.rb

@@ -0,0 +1 @@
+require 'lib/spamham.rb'

data/lib/spamham/configure.rb

@@ -0,0 +1,25 @@
+class SpamHam
+	def conf(config = nil)
+		@config = config || {}
+		load_config_from_defaults
+		if File.exist?("config/spamham.rb")
+			load_config_from_file
+		end
+	end
+
+	private
+	
+	def load_config_from_file
+		load 'config/spamham.rb'
+		user_config
+	end
+
+	#Sets and unset values to thier defaults
+	def load_config_from_defaults
+		@config[:trigger_weight] ||= 30
+		@config[:link_weight] ||= 10
+		@config[:link_multiple] ||= 1.1
+		@config[:multiple_link_style_weight] ||= 20
+		@config[:link_styles] ||= [/<a href="(.*?)">(.*?)<\/a>/,/\[url=(.*?)\](.*?)\[\/url\]/]
+	end
+end

data/lib/spamham/link.rb

@@ -0,0 +1,21 @@
+class SpamHam
+	class Link
+		def initialize(string)
+			@host="s"
+			@href="http://"
+			@message="s"
+		end
+
+		def host
+			@host
+		end
+
+		def href
+			@href
+		end
+
+		def message
+			@message
+		end
+	end
+end

data/lib/spamham/scan.rb

@@ -0,0 +1,32 @@
+class SpamHam
+	def scan
+		@string_weight = 0
+		build_link_weights
+		build_word_list_weights
+	end
+
+	private
+
+	def build_link_weights
+		links.each do |linkstyle|
+			linkstyle.each do |link|
+				@string_weight = @string_weight + (@config[:link_weight] * (@config[:link_multiple] * (linkstyle.index(link) + 1)))
+			end
+		end
+		@string_weight = @string_weight + @config[:multiple_link_style_weight] if links.size >= 2
+	end
+	
+	def links
+		matched_links = []
+		@config[:link_styles].each do |style|
+			matched_links.push(@string.scan(style)) unless @string.scan(style) == [] and matched_links.index(@string.scan(style)) != nil
+		end
+		matched_links
+	end
+
+	def build_word_list_weights
+		wordlist.each do |word|
+			@string_weight = @string_weight + (@string.scan(/#{word}/).count)
+		end
+	end
+end

data/lib/spamham/string.rb

@@ -0,0 +1,10 @@
+class String
+	# Adds .spam? to strings
+	#
+	# Takes configuration options
+	def spam?(config = nil)
+		spam=SpamHam.new
+		spam.conf(config) if config
+		spam.spam?(self)
+	end
+end

data/lib/spamham/version.rb

@@ -0,0 +1,3 @@
+class SpamHam
+  VERSION = "0.0.2"
+end

data/lib/spamham/wordlist.rb

@@ -0,0 +1,18 @@
+class SpamHam
+	def wordlist
+		[
+			"viagra",
+			"fake watches",
+			"enlargement",
+			"penis",
+			"designer",
+			"brands",
+			"buy",
+			"online slots",
+			"valium",
+			"ambien",
+			"dosage",
+			"propecia"
+		] + (@config[:additional_words] || [])
+	end
+end

data/lib/spamham.rb

@@ -0,0 +1,21 @@
+require 'spamham/configure.rb'
+require 'spamham/scan.rb'
+require 'spamham/string.rb'
+require 'spamham/wordlist.rb'
+
+class SpamHam
+	def initialize
+		self.conf
+	end
+
+	def spam?(string)
+		@string=string
+		self.scan
+		@string=nil
+		@string_weight >= @config[:trigger_weight]
+	end
+	
+	def weight
+		@string_weight
+	end
+end

data/spamham.gemspec

@@ -0,0 +1,21 @@
+# -*- encoding: utf-8 -*-
+$:.push File.expand_path("../lib", __FILE__)
+require "spamham/version"
+
+Gem::Specification.new do |s|
+  s.name        = "spamham"
+  s.version     = SpamHam::VERSION
+  s.platform    = Gem::Platform::RUBY
+  s.authors     = ["Adam \"Arcath\" Laycock"]
+  s.email       = ["adam@arcath.net"]
+  s.homepage    = "http://www.arcath.net"
+  s.summary     = %q{Provides Spam filtering for ruby and rails}
+  s.description = %q{...}
+
+  s.rubyforge_project = "spamham"
+
+  s.files         = `git ls-files`.split("\n")
+  s.test_files    = `git ls-files -- {test,spec,features}/*`.split("\n")
+  s.executables   = `git ls-files -- bin/*`.split("\n").map{ |f| File.basename(f) }
+  s.require_paths = ["lib"]
+end

data/spec/spamham_spec.rb

@@ -0,0 +1,35 @@
+require 'lib/spamham.rb'
+
+describe SpamHam, "#spam?" do
+	it "Should return false for \"not spam\"" do
+		SpamHam.new.spam?("not spam").should == false
+	end
+
+	it "Should return true for \"<a href=\"host\">text</a> [url=host1]text1[/url] \"<a href=\"host2\">text2</a>\"" do
+		SpamHam.new.spam?("<a href=\"host\">text</a> [url=host1]text1[/url] \"<a href=\"host2\">text2</a>").should == true
+	end
+
+	it "Should return true for \"ixxdzakarvtstdldkpxt, [url=http://www.amosboergoats.com/]Buy Cialis[/url], hgOvsFa.\"" do
+		ham = SpamHam.new.spam?("ixxdzakarvtstdldkpxt, [url=http://www.amosboergoats.com/]Buy Cialis[/url], hgOvsFa.").should == true
+	end
+end
+
+describe SpamHam, "#weight" do
+	it "Should return 1 for \"penis\"" do
+		ham=SpamHam.new
+		ham.spam?("penis")
+		ham.weight.should == 1
+	end
+
+	it "Should return 3.6 for \"designer brands\"" do
+		ham=SpamHam.new
+		ham.spam?("designer brands")
+		ham.weight.should <= 3.7
+	end
+end
+
+describe String, "#spam?" do
+	it "Should return flase for \"string of characters\"" do
+		"string of characters".spam?.should == false
+	end
+end

metadata

@@ -0,0 +1,81 @@
+--- !ruby/object:Gem::Specification 
+name: spamham
+version: !ruby/object:Gem::Version 
+  hash: 27
+  prerelease: false
+  segments: 
+  - 0
+  - 0
+  - 2
+  version: 0.0.2
+platform: ruby
+authors: 
+- Adam "Arcath" Laycock
+autorequire: 
+bindir: bin
+cert_chain: []
+
+date: 2010-12-23 00:00:00 +00:00
+default_executable: 
+dependencies: []
+
+description: ...
+email: 
+- adam@arcath.net
+executables: []
+
+extensions: []
+
+extra_rdoc_files: []
+
+files: 
+- .gitignore
+- Gemfile
+- README.rdoc
+- Rakefile
+- init.rb
+- lib/spamham.rb
+- lib/spamham/configure.rb
+- lib/spamham/link.rb
+- lib/spamham/scan.rb
+- lib/spamham/string.rb
+- lib/spamham/version.rb
+- lib/spamham/wordlist.rb
+- spamham.gemspec
+- spec/spamham_spec.rb
+has_rdoc: true
+homepage: http://www.arcath.net
+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: spamham
+rubygems_version: 1.3.7
+signing_key: 
+specification_version: 3
+summary: Provides Spam filtering for ruby and rails
+test_files: []
+