gman 4.0.0 → 4.1.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 63b98f8e56e706ce94ce5304d04be1cceb9f83a7
-  data.tar.gz: 7223ef35fb6eeb25c3e162f27ce90169caf57413
+  metadata.gz: 581df932b6e1afe6a5ed61c5116abee789b3da65
+  data.tar.gz: b7bb194063df9af7d3b3a3d83b287c94d9076170
 SHA512:
-  metadata.gz: 49c5abd646e7415d1f704bf25d2cab4910410c3d036c72f8cfd238e75a82ad1ceb6eab91cbcb0d7a5426c08c72be6c70aae04a8528c790fbc26291d7caaa409c
-  data.tar.gz: d50a8c56755e0c056ec6c358d9daa76928283a2f079f992cd5027dbbf9851c0de23b62ca37d81c99805f79670676bd273292ab0ff408235f97db1faf2422d0ce
+  metadata.gz: 59cefe32f4435eb85a7d8621206314328bd9804c9a63eeef6bcf68125ae6bcd4f5ed1fc237b1925d9e198ee1ae4332a3f0d381d91761bb6a438cdf5e781f6e56
+  data.tar.gz: 899fe3d0dc4c9d5cfd75ab55bf64e08220cdf112140e9cb573c0009517ef3e6689a31fe5587b9209e7d26538cab88d946e58b183ec0de0495384cbebf13a3977

data/lib/gman.rb

@@ -1,114 +1,11 @@
-require 'public_suffix'
-require 'yaml'
+require 'naughty_or_nice'
 require 'swot'
-require "addressable/uri"
 require 'iso_country_codes'
-require_relative "gman/version"
-
-class Gman
-
-  # Source: http://bit.ly/1n2X9iv
-  EMAIL_REGEX = %r{
-        ^
-        (
-          [\w\!\#\$\%\&\'\*\+\-\/\=\?\^\`\{\|\}\~]+
-          \.
-        )
-        *
-        [\w\!\#\$\%\&\'\*\+\-\/\=\?\^\`\{\|\}\~]+
-        @
-        (
-          (
-            (
-              (
-                (
-                  [a-z0-9]{1}
-                  [a-z0-9\-]{0,62}
-                  [a-z0-9]{1}
-                )
-                |
-                [a-z]
-              )
-              \.
-            )+
-            [a-z]{2,6}
-          )
-          |
-          (
-            \d{1,3}
-            \.
-          ){3}
-          \d{1,3}
-          (
-            \:\d{1,5}
-          )?
-        )
-        $
-      }xi
-
-  # Second level .us domains for states and locality
-  # See http://en.wikipedia.org/wiki/.us
-  #
-  # Examples:
-  #  * foo.state.il.us
-  #  * ci.foo.il.us
-  #
-  # Not:
-  #  * state.foo.il.us
-  #  * foo.ci.il.us
-  #  * k12.il.us
-  #  * ci.foo.zx.us
-  LOCALITY_REGEX = %r{
-    (
-      (state|dst|cog)
-    |
-      (ci|town|vil|co)\.[a-z-]+
-    )
-    \.(ak|al|ar|az|ca|co|ct|dc|de|fl|ga|hi|ia|id|il|in|ks|ky|la|ma|md|me|mi|mn|mo|ms|mt|nc|nd|ne|nh|nj|nm|nv|ny|oh|ok|or|pa|ri|sc|sd|tn|tx|um|ut|va|vt|wa|wi|wv|wy)
-    \.us
-     }x
-
-  # Map last part of TLD to alpha2 country code
-  ALPHA2_MAP = {
-    :ac     => 'sh',
-    :uk     => 'gb',
-    :su     => 'ru',
-    :tp     => 'tl',
-    :yu     => 'rs',
-    :gov    => "us",
-    :mil    => "us",
-    :org    => "us",
-    :com    => "us",
-    :net    => "us",
-    :edu    => "us",
-    :travel => "us",
-    :info   => "us"
-  }
+require_relative 'gman/country_codes'
+require_relative 'gman/locality'
 
+class Gman < NaughtyOrNice
   class << self
-
-    attr_writer :list
-
-    # Normalizes and checks if a given string represents a government domain
-    # Possible strings to test:
-    #   ".gov"
-    #   "foo.gov"
-    #   "foo@bar.gov"
-    #   "foo.gov.uk"
-    #   "http://foo.bar.gov"
-    #
-    # Returns boolean true if a government domain
-    def valid?(text)
-      Gman.new(text).valid?
-    end
-
-    # Is the given string in the form of a valid email address?
-    #
-    # Returns true if email, otherwise false
-    def email?(text)
-      Gman.new(text).email?
-    end
-
     # returns an instance of our custom public suffix list
     # list behaves like PublicSuffix::List but is limited to our whitelisted domains
     def list
@@ -117,46 +14,10 @@ class Gman
 
     # Returns the absolute path to the domain list
     def list_path
-      File.join(File.dirname(__FILE__), "domains.txt")
+      File.join(File.dirname(__FILE__), "../config/domains.txt")
     end
   end
 
-  # Creates a new Gman instance
-  #
-  # text - the input string to check for governmentiness
-  def initialize(text)
-    @text = text.to_s.downcase.strip
-  end
-
-  # Parse the domain from the input string
-  #
-  # Can handle urls, domains, or emails
-  #
-  # Returns the domain string
-  def domain
-    @domain ||= begin
-      return nil if @text.empty?
-
-      uri = Addressable::URI.parse(@text)
-
-      if uri.host # valid https?://* URI
-        uri.host
-      elsif email?
-        @text.match(/@([\w\.\-]+)\Z/i)[1]
-      else # url sans http://
-        begin
-          uri = Addressable::URI.parse("http://#{@text}")
-          # properly parse http://foo edge cases
-          # see https://github.com/sporkmonger/addressable/issues/145
-          uri.host if uri.host =~ /\./
-        rescue Addressable::URI::InvalidURIError
-          nil
-        end
-      end
-    end
-  end
-  alias_method :to_s, :domain
-
   # Checks if the input string represents a government domain
   #
   # Returns boolean true if a government domain
@@ -177,54 +38,4 @@ class Gman
     # also allow for explicit matches to domain list
     Gman.list.rules.any? { |rule| rule.value == domain }
   end
-
-  # Is the input text in the form of a valid email address?
-  #
-  # Returns true if email, otherwise false
-  def email?
-    !!(@text =~ EMAIL_REGEX)
-  end
-
-  # Is this domain a .us state or locality?
-  def locality?
-    !!(domain =~ LOCALITY_REGEX)
-  end
-
-  # Helper function to return the public suffix domain object
-  #
-  # Supports all domain strings (URLs, emails)
-  #
-  # Returns the domain object or nil, but no errors, never an error
-  def domain_parts
-    PublicSuffix.parse domain
-  rescue PublicSuffix::DomainInvalid
-    nil
-  end
-
-  # Returns the two character alpha county code represented by the domain
-  #
-  # e.g., United States = US, United Kingdom = GB
-  def alpha2
-    alpha2 = domain_parts.tld.split('.').last
-    if ALPHA2_MAP[alpha2.to_sym]
-      ALPHA2_MAP[alpha2.to_sym]
-    else
-      alpha2
-    end
-  end
-
-  # Returns the ISO Country represented by the domain
-  #
-  # Example Usage:
-  # Gman.new("foo.gov").country.name     => "United States"
-  # Gman.new("foo.gov").country.currency => "USD"
-  def country
-    @country ||= IsoCountryCodes.find(alpha2)
-  end
-
-  # Console output
-  def inspect
-    "#<Gman domain=\"#{domain}\" valid=#{valid?}>"
-  end
-
 end

data/lib/gman/country_codes.rb

@@ -0,0 +1,40 @@
+class Gman < NaughtyOrNice
+
+  # Map last part of TLD to alpha2 country code
+  ALPHA2_MAP = {
+    :ac     => 'sh',
+    :uk     => 'gb',
+    :su     => 'ru',
+    :tp     => 'tl',
+    :yu     => 'rs',
+    :gov    => "us",
+    :mil    => "us",
+    :org    => "us",
+    :com    => "us",
+    :net    => "us",
+    :edu    => "us",
+    :travel => "us",
+    :info   => "us"
+  }
+
+  # Returns the two character alpha county code represented by the domain
+  #
+  # e.g., United States = US, United Kingdom = GB
+  def alpha2
+    alpha2 = domain_parts.tld.split('.').last
+    if ALPHA2_MAP[alpha2.to_sym]
+      ALPHA2_MAP[alpha2.to_sym]
+    else
+      alpha2
+    end
+  end
+
+  # Returns the ISO Country represented by the domain
+  #
+  # Example Usage:
+  # Gman.new("foo.gov").country.name     => "United States"
+  # Gman.new("foo.gov").country.currency => "USD"
+  def country
+    @country ||= IsoCountryCodes.find(alpha2)
+  end
+end

data/lib/gman/locality.rb

@@ -0,0 +1,28 @@
+class Gman < NaughtyOrNice
+
+  LOCALITY_REGEX = %r{
+    (
+      (state|dst|cog)
+    |
+      (ci|town|vil|co)\.[a-z-]+
+    )
+    \.(ak|al|ar|az|ca|co|ct|dc|de|fl|ga|hi|ia|id|il|in|ks|ky|la|ma|md|me|mi|mn|mo|ms|mt|nc|nd|ne|nh|nj|nm|nv|ny|oh|ok|or|pa|ri|sc|sd|tn|tx|um|ut|va|vt|wa|wi|wv|wy)
+    \.us
+     }x
+
+  # Second level .us domains for states and locality
+  # See http://en.wikipedia.org/wiki/.us
+  #
+  # Examples:
+  #  * foo.state.il.us
+  #  * ci.foo.il.us
+  #
+  # Not:
+  #  * state.foo.il.us
+  #  * foo.ci.il.us
+  #  * k12.il.us
+  #  * ci.foo.zx.us
+  def locality?
+    !!(domain =~ LOCALITY_REGEX)
+  end
+end

metadata

@@ -1,73 +1,73 @@
 --- !ruby/object:Gem::Specification
 name: gman
 version: !ruby/object:Gem::Version
-  version: 4.0.0
+  version: 4.1.0
 platform: ruby
 authors:
 - Ben Balter
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2014-09-18 00:00:00.000000000 Z
+date: 2014-09-19 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
-  name: public_suffix
+  name: swot
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '1.4'
+        version: 0.3.1
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '1.4'
+        version: 0.3.1
 - !ruby/object:Gem::Dependency
-  name: swot
+  name: iso_country_codes
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 0.3.1
+        version: '0.4'
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 0.3.1
+        version: '0.4'
 - !ruby/object:Gem::Dependency
-  name: addressable
+  name: naughty_or_nice
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '2.3'
+        version: 0.0.2
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '2.3'
+        version: 0.0.2
 - !ruby/object:Gem::Dependency
-  name: iso_country_codes
+  name: rake
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - "~>"
+    - - ">="
       - !ruby/object:Gem::Version
-        version: '0.4'
-  type: :runtime
+        version: '0'
+  type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - "~>"
+    - - ">="
       - !ruby/object:Gem::Version
-        version: '0.4'
+        version: '0'
 - !ruby/object:Gem::Dependency
-  name: rake
+  name: shoulda
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - ">="
@@ -81,7 +81,7 @@ dependencies:
       - !ruby/object:Gem::Version
         version: '0'
 - !ruby/object:Gem::Dependency
-  name: shoulda
+  name: rdoc
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - ">="
@@ -95,7 +95,7 @@ dependencies:
       - !ruby/object:Gem::Version
         version: '0'
 - !ruby/object:Gem::Dependency
-  name: rdoc
+  name: bundler
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - ">="
@@ -109,7 +109,7 @@ dependencies:
       - !ruby/object:Gem::Version
         version: '0'
 - !ruby/object:Gem::Dependency
-  name: bundler
+  name: pry
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - ">="
@@ -130,14 +130,12 @@ executables:
 extensions: []
 extra_rdoc_files: []
 files:
-- CONTRIBUTING.md
-- Gemfile
 - LICENSE
-- README.md
 - bin/gman_filter
-- lib/domains.txt
+- config/domains.txt
 - lib/gman.rb
-- lib/gman/version.rb
+- lib/gman/country_codes.rb
+- lib/gman/locality.rb
 homepage: https://github.com/benbalter/gman
 licenses:
 - MIT

data/CONTRIBUTING.md

@@ -1,22 +0,0 @@
-# Contributing to Gman
-
-## How to contribute
-
-1. Fork the project
-2. Create a descriptive branch
-3. Make your change
-4. Submit a pull request
-
-## Code
-
-Open an issue, or submit a pull request
-
-## Domains
-
-Domains live in `./lib/domains.txt` as a list of TLDs and SLD+TLDs.
-
-Right now, the only valid government top level domains (TLDs), represent the US government and are `.gov`, and `.mil`.
-
-Secondary domains (e.g., `gov.uk`, or `mil.au`) detect non-US government entities.
-
-To add or remove a domain from the list of known government domains, simply edit the `domains.txt` file.

data/Gemfile

@@ -1,6 +0,0 @@
-source 'https://rubygems.org'
-gemspec
-
-group :development, :test do
-  gem 'net-dns'
-end

data/README.md

@@ -1,68 +0,0 @@
-# Gman Gem
-
-[![Build Status](https://travis-ci.org/benbalter/gman.png)](https://travis-ci.org/benbalter/gman) [![Gem Version](https://badge.fury.io/rb/gman.png)](http://badge.fury.io/rb/gman)
-
-A ruby gem to check if the owner of a given email address or website is working for THE MAN (a.k.a verifies government domains). It does this by leveraging the power of the [Public Suffix List](http://publicsuffix.org/), and the associated [Ruby Gem](https://github.com/weppos/publicsuffix-ruby).
-
-You could theoretically [use regex](https://gist.github.com/benbalter/6147066), but either you'll get a bunch of false positives, or your regex will be insanely complicated. `gov.uk`, may be valid, for example, but `gov.fr` is not (it's `gouv.fr`, for what it's worth).
-
-The solution? Use Public Suffix to verify that it's a valid public domain, then maintain [a crowd-sourced sub-list of known global government and military domains](https://github.com/benbalter/gman/blob/master/lib/domains.txt). It should cover all US and international, government and military domains for both email and website verification.
-
-See a domains that's missing or one that shouldn't be there? [We'd love you to contribute](CONTRIBUTING.md).
-
-## Installation
-
-Gman is a Ruby gem, so you'll need a little Ruby-fu to get it working. Simply
-
-`gem install gman`
-
-Or add this to your `Gemfile` before doing a `bundle install`:
-
-`gem 'gman'`
-
-## Usage
-
-### In general
-
-### Verify email addresses
-
-```ruby
-Gman.valid? "foo@bar.gov" #=> true
-Gman.valid? "foo@bar.com" #=> false
-```
-
-### Verify domain
-
-```ruby
-Gman.valid? "http://foo.bar.gov" #=> true
-Gman.valid? "foo.bar.gov"        #=> true
-Gman.valid? "foo.gov"            #=> true
-Gman.valid? "foo.biz"            #=> false
-```
-
-### Get the ISO Country Code information represented by a government domain
-
-```ruby
-domain = Gman.new "whitehouse.gov" #=> #<Gman domain="whitehouse.gov" valid=true>
-domain.country.name                #=> "United States"
-domain.country.alpha2              #=> "US"
-domain.country.alpha3              #=> "USA"
-domain.country.currency            #=> "USD"
-domain.conutry.calling_code        #=> "+1"
-```
-
-### Command line
-
-Filters newline-separated email addresses from stdin. Example usage:
-
-```
-$ gman_filter < path/to/list/of/addresses.txt
-```
-
-## Contributing
-
-Contributions welcome! Please see [the contribution guidelines](CONTRIBUTING.md) for code contributions or for details on how to add, update, or delete government domains.
-
-## Credits
-
-Heavily inspired by [swot](https://github.com/leereilly/swot). Thanks [@leereilly](https://github.com/leereilly)!

data/lib/gman/version.rb

@@ -1,3 +0,0 @@
-class Gman
-  VERSION = '4.0.0'
-end