cf-swearjar 1.0.1

data/lib/swearjar/tester.rb

@@ -0,0 +1,32 @@
+require 'yaml'
+require 'fuzzy_hash'
+require 'bloomfilter'
+
+class Swearjar
+  class Tester
+
+    def initialize(config_file)
+      data = YAML.load_file
+
+      @tester = FuzzyHash.new
+
+      data['regex'].each do |pattern, type|
+        @tester[Regexp.new(pattern)] = type
+      end
+
+      data['simple'].each do |test, type|
+        @tester[test] = type
+      end
+
+    end
+
+    def scan(string, &block)
+      string.scan(/\b[\b]+\b/, &block)
+    end
+
+    def profane?(string)
+      scan(string) {|w| return true}
+    end
+
+  end
+end

data/lib/swearjar/version.rb

@@ -0,0 +1,3 @@
+class Swearjar
+  VERSION = '1.0.1'
+end

data/lib/swearjar.rb

@@ -0,0 +1,63 @@
+require 'yaml'
+require 'fuzzy_hash'
+
+class Swearjar
+
+  def self.default
+    from_language
+  end
+
+  def self.from_language(language = 'en')
+    new(File.join(File.dirname(__FILE__), 'config', "#{language}.yml"))
+  end
+
+  attr_reader :tester, :hash
+
+  def initialize(file = nil)
+    @tester = FuzzyHash.new
+    @hash = {}
+    load_file(file) if file
+  end
+
+  def load_file(file)
+    data = YAML.load_file(file)
+
+    data['regex'].each do |pattern, type|
+      @tester[Regexp.new(pattern)] = type
+    end if data['regex']
+
+    data['simple'].each do |test, type|
+      @hash[test] = type
+    end if data['simple']
+  end
+
+  def scan(string, &block)
+    string = string.to_s
+    string.scan(/\b[a-zA-Z-]+\b/) do |word|
+      block.call(word, hash[word.downcase] || hash[word.downcase.gsub(/e?s$/,'')] )
+    end
+    if match = tester.match_with_result(string)
+      block.call(match.last, match.first)
+    end
+  end
+
+  def profane?(string)
+    string = string.to_s
+    scan(string) {|word, test| return true if !test.nil?}
+    return false
+  end
+
+  def scorecard(string)
+    string = string.to_s
+    scorecard = {}
+    scan(string) {|word, test| test.each { |type| scorecard.key?(type) ? scorecard[type] += 1 : scorecard[type] = 1} if test}
+    scorecard
+  end
+
+  def censor(string)
+    censored_string = string.to_s.dup
+    scan(string) {|word, test| censored_string.gsub!(word, block_given? ? yield(word) : word.gsub(/\S/, '*')) if test}
+    censored_string
+  end
+
+end

data/spec/data/swear.yml

@@ -0,0 +1,2 @@
+simple:
+  "python": ["sexual"]

data/spec/spec.opts

@@ -0,0 +1,7 @@
+--colour
+--format
+specdoc
+--loadby
+mtime
+--reverse
+--backtrace

data/spec/spec_helper.rb

@@ -0,0 +1 @@
+require File.expand_path(File.join(File.dirname(__FILE__), '..', 'lib', 'swearjar'))

data/spec/swearjar_spec.rb

@@ -0,0 +1,48 @@
+require 'spec_helper'
+
+describe Swearjar do
+
+  it "should detect dirty words" do
+    Swearjar.default.profane?('fuck you jim henson').should be_true
+  end
+
+  it "should detect dirty words regardless of case" do
+    Swearjar.default.profane?('FuCk you jim henson').should be_true
+  end
+
+  it "should not detect non-dirty words" do
+    Swearjar.default.profane?('i love you jim henson').should be_false
+  end
+
+  it "should give us a scorecard" do
+    Swearjar.default.scorecard('fuck you jim henson').should == {'sexual'=>1}
+  end
+
+  it "should detect multiword" do
+    Swearjar.default.scorecard('jim henson has a hard on').should == {'sexual'=>1}
+  end
+
+  it "should detect multiword plurals" do
+    Swearjar.default.scorecard('jim henson has a hard ons').should == {'sexual'=>1}
+  end
+
+  it "should detect simple dirty plurals" do
+    Swearjar.default.profane?('jim henson had two dicks').should be_true
+    Swearjar.default.profane?('jim henson has two asses').should be_true
+  end
+
+  it "should censor a string" do
+    Swearjar.default.censor('jim henson has a massive hard on he is gonna use to fuck everybody').should == 'jim henson has a massive **** ** he is gonna use to **** everybody'
+  end
+
+  it "should not do much when given a non-string" do
+    Swearjar.default.profane?(nil).should be_false
+  end
+
+  it "should allow you to load a new yaml file" do
+    sj = Swearjar.new
+    sj.load_file(File.expand_path('../data/swear.yml', __FILE__))
+    sj.censor("Python is the best language!").should == "****** is the best language!"
+  end
+
+end

metadata

@@ -0,0 +1,113 @@
+--- !ruby/object:Gem::Specification
+name: cf-swearjar
+version: !ruby/object:Gem::Version
+  version: 1.0.1
+  prerelease: 
+platform: ruby
+authors:
+- Joshua Hull
+autorequire: 
+bindir: bin
+cert_chain: []
+date: 2012-06-08 00:00:00.000000000 Z
+dependencies:
+- !ruby/object:Gem::Dependency
+  name: fuzzyhash
+  requirement: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: 0.0.11
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: 0.0.11
+- !ruby/object:Gem::Dependency
+  name: rake
+  requirement: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: 0.8.7
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: 0.8.7
+- !ruby/object:Gem::Dependency
+  name: rspec
+  requirement: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: 1.3.0
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: 1.3.0
+description: Put another nickel in the swearjar. Simple profanity detection with content
+  analysis.
+email: joshbuddy@gmail.com
+executables: []
+extensions: []
+extra_rdoc_files:
+- README.rdoc
+files:
+- .gitignore
+- Gemfile
+- README.rdoc
+- Rakefile
+- cf-swearjar.gemspec
+- lib/config/en.yml
+- lib/swearjar.rb
+- lib/swearjar/tester.rb
+- lib/swearjar/version.rb
+- spec/data/swear.yml
+- spec/spec.opts
+- spec/spec_helper.rb
+- spec/swearjar_spec.rb
+homepage: http://github.com/joshbuddy/swearjar
+licenses: []
+post_install_message: 
+rdoc_options:
+- --charset=UTF-8
+require_paths:
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement
+  none: false
+  requirements:
+  - - ! '>='
+    - !ruby/object:Gem::Version
+      version: '0'
+required_rubygems_version: !ruby/object:Gem::Requirement
+  none: false
+  requirements:
+  - - ! '>='
+    - !ruby/object:Gem::Version
+      version: '0'
+requirements: []
+rubyforge_project: swearjar
+rubygems_version: 1.8.24
+signing_key: 
+specification_version: 3
+summary: Put another nickel in the swearjar. Simple profanity detection with content
+  analysis
+test_files:
+- spec/data/swear.yml
+- spec/spec.opts
+- spec/spec_helper.rb
+- spec/swearjar_spec.rb