si-copyscape 0.1.0

checksums.yaml

@@ -0,0 +1,7 @@
+---
+SHA1:
+  metadata.gz: 6aa16fb6129dc878e9954870974c18f3e1e6f6ee
+  data.tar.gz: 94caa08afc29b45beeabb904e83ad3b3b9261e5e
+SHA512:
+  metadata.gz: d9aa789ecab6eba516c76ca4da9ca91eb5d85a05eefb1611513aab46ecdee978d5d94054b82b27b845e5dd1f7c53c626bc673e4a45a3b460ce920a4ef72a102c
+  data.tar.gz: 49abd6235ea1b75ff648a4c82de9b6c29fe4513ef56e461a4052dca808be675e8cc8d12cb3c9a8cadecaeb021f0d87ebe2e676f08d90e2dc9a9cd4ebf62d4213

data/.gitignore

@@ -0,0 +1 @@
+*.gem

data/Gemfile

@@ -0,0 +1,3 @@
+source 'https://rubygems.org'
+
+gemspec

data/Gemfile.lock

@@ -0,0 +1,64 @@
+PATH
+  remote: .
+  specs:
+    si-copyscape (0.1.0)
+      crack
+
+GEM
+  remote: https://rubygems.org/
+  specs:
+    addressable (2.3.8)
+    coderay (1.1.0)
+    crack (0.4.2)
+      safe_yaml (~> 1.0.0)
+    ffi (1.9.10)
+    formatador (0.2.5)
+    guard (2.13.0)
+      formatador (>= 0.2.4)
+      listen (>= 2.7, <= 4.0)
+      lumberjack (~> 1.0)
+      nenv (~> 0.1)
+      notiffany (~> 0.0)
+      pry (>= 0.9.12)
+      shellany (~> 0.0)
+      thor (>= 0.18.1)
+    guard-compat (1.2.1)
+    guard-minitest (2.4.4)
+      guard-compat (~> 1.2)
+      minitest (>= 3.0)
+    listen (3.0.3)
+      rb-fsevent (>= 0.9.3)
+      rb-inotify (>= 0.9)
+    lumberjack (1.0.9)
+    method_source (0.8.2)
+    minitest (5.8.0)
+    nenv (0.2.0)
+    notiffany (0.0.7)
+      nenv (~> 0.1)
+      shellany (~> 0.0)
+    pry (0.10.1)
+      coderay (~> 1.1.0)
+      method_source (~> 0.8.1)
+      slop (~> 3.4)
+    rb-fsevent (0.9.5)
+    rb-inotify (0.9.5)
+      ffi (>= 0.5.0)
+    safe_yaml (1.0.4)
+    shellany (0.0.1)
+    slop (3.6.0)
+    thor (0.19.1)
+    vcr (2.9.3)
+    webmock (1.21.0)
+      addressable (>= 2.3.6)
+      crack (>= 0.3.2)
+
+PLATFORMS
+  ruby
+
+DEPENDENCIES
+  guard
+  guard-minitest
+  minitest
+  si-copyscape!
+  vcr
+  webmock

data/Guardfile

@@ -0,0 +1,42 @@
+# A sample Guardfile
+# More info at https://github.com/guard/guard#readme
+
+## Uncomment and set this to only include directories you want to watch
+# directories %w(app lib config test spec features) \
+#  .select{|d| Dir.exists?(d) ? d : UI.warning("Directory #{d} does not exist")}
+
+## Note: if you are using the `directories` clause above and you are not
+## watching the project directory ('.'), then you will want to move
+## the Guardfile to a watched dir and symlink it back, e.g.
+#
+#  $ mkdir config
+#  $ mv Guardfile config/
+#  $ ln -s config/Guardfile .
+#
+# and, you'll have to watch "config/Guardfile" instead of "Guardfile"
+
+guard :minitest do
+  # with Minitest::Unit
+  watch(%r{^test/(.*)\/?test_(.*)\.rb$})
+  watch(%r{^lib/(.*/)?([^/]+)\.rb$})     { |m| "test/#{m[1]}test_#{m[2]}.rb" }
+  watch(%r{^test/test_helper\.rb$})      { 'test' }
+
+  # with Minitest::Spec
+  # watch(%r{^spec/(.*)_spec\.rb$})
+  # watch(%r{^lib/(.+)\.rb$})         { |m| "spec/#{m[1]}_spec.rb" }
+  # watch(%r{^spec/spec_helper\.rb$}) { 'spec' }
+
+  # Rails 4
+  # watch(%r{^app/(.+)\.rb$})                               { |m| "test/#{m[1]}_test.rb" }
+  # watch(%r{^app/controllers/application_controller\.rb$}) { 'test/controllers' }
+  # watch(%r{^app/controllers/(.+)_controller\.rb$})        { |m| "test/integration/#{m[1]}_test.rb" }
+  # watch(%r{^app/views/(.+)_mailer/.+})                   { |m| "test/mailers/#{m[1]}_mailer_test.rb" }
+  # watch(%r{^lib/(.+)\.rb$})                               { |m| "test/lib/#{m[1]}_test.rb" }
+  # watch(%r{^test/.+_test\.rb$})
+  # watch(%r{^test/test_helper\.rb$}) { 'test' }
+
+  # Rails < 4
+  # watch(%r{^app/controllers/(.*)\.rb$}) { |m| "test/functional/#{m[1]}_test.rb" }
+  # watch(%r{^app/helpers/(.*)\.rb$})     { |m| "test/helpers/#{m[1]}_test.rb" }
+  # watch(%r{^app/models/(.*)\.rb$})      { |m| "test/unit/#{m[1]}_test.rb" }
+end

data/README.md

@@ -0,0 +1,112 @@
+# SI::CopyScape Gem
+A gem to provide the communication layer with the Copyscape.com Premium API and your application.
+
+Wiki: (requires login) http://wiki.searchinfluence.com/index.php?title=CopyScape
+
+API Documentation: (requires login) http://copyscape.com/apiconfigure.php
+
+## Usage:
+Set the following environment variables (or pass the values in during initialization):
+- ENV['COPYSCAPE_USERNAME']
+- ENV['COPYSCAPE_API_KEY']
+
+Find the most recent version of this gem at http://gems.searchinfluence.com/gems/si-copyscape and then specify the version number when adding this gem to your application's Gemfile.
+
+```ruby
+# Add the following to your Gemfile
+source 'http://user:pass@gems.searchinfluence.com' do
+  gem 'si-copyscape', '0.0.0'
+end
+```
+
+Instantiate the copyscape object with the following optional paramaters
+```ruby
+copyscape = SI::CopyScape.new(
+  username: "foo", # not required if ENV['COPYSCAPE_USERNAME'] is set
+  api_key: "bar", # not required if ENV['COPYSCAPE_API_KEY'] is set
+  uri: "http://foo", # defaults to http://www.copyscape.com/api/
+  match_percent: 10 # defaults to 5
+)
+```
+
+Get Remaining Credit Information
+```ruby
+copyscape.credit_balance
+#=><struct SI::CopyScape::Balance value=112.62, total=2252, today=2252>
+```
+
+Perform text searches for copy matches
+```ruby
+text = "A national website promotion company, Search Influence routinely delivers a 10:1 return on investment, or better, for our customers."
+
+# Performs a public internet search (CREDIT COST: 1)
+copyscape.internet_matches! text
+
+# Performs a private index search (CREDIT COST: 1)
+copyscape.private_matches! text
+
+# Performs a public internet & private index search (CREDIT COST: 2)
+copyscape.internet_and_private_matches! text
+
+# All these methods return an array of data structs ex:
+#=>
+# [
+#  <struct SI::CopyScape::Match
+#        words_matched = 20,
+#      percent_matched = 100,
+#                title = "Search Influence | Website Promotion Company",
+#                  url = "http://www.searchinfluence.com/",
+#        copyscape_url = "http://view.copyscape.com/compare/wpbdhatumu/1",
+#         text_snippet = "... Trusted, Scalable Search, Social and Online Advertising. A national website promotion company, Search Influence routinely delivers a 10:1 return on investment, or better, for our customers.",
+#         html_snippet = "<font color=\"#777777\">... Trusted, Scalable Search, Social and Online Advertising. </font><font color=\"#000000\">A national website promotion company, Search Influence routinely delivers a 10:1 return on investment, or better, for our customers.</font>"
+#  >
+# ]
+```
+
+Add text to our private index on Copyscape.com
+```ruby
+copyscape.add_to_private_index(
+ text: 'Text to add to index',
+ title: 'Title', # not required
+ id: 420 # not required
+)
+#=><struct SI::CopyScape::PrivateIndex words=5, handle="SIA_2_E00JOQ0A2W_T1Q2J78LA1", id="420", title="Title">
+```
+
+If there is an error, you can get a string describing the error (returns nil if there is no error)
+```ruby
+copyscape.internet_matches! "test"
+copyscape.error
+#=>"At least 15 words are required to perform a search"
+```
+
+The error method is also available on the SI::CopyscapeMatches collection returned after a search
+```ruby
+matches = copyscape.internet_matches! "test"
+matches.error
+#=>"At least 15 words are required to perform a search"
+```
+
+## Flog Score
+```
+   177.7: flog total
+     5.2: flog/method average
+
+    35.0: SI::CopyscapeMatches total
+    16.5: SI::CopyscapeMatches#_build_collection lib/copyscape_matches.rb:28
+    11.0: SI::CopyscapeMatches#_without_rejects lib/copyscape_matches.rb:47
+     7.5: SI::CopyscapeMatches#initialize  lib/copyscape_matches.rb:7
+
+    31.6: SI::CopyScape total
+    11.0: SI::CopyScape#credit_balance     lib/copyscape.rb:23
+     9.6: SI::CopyScape#add_to_private_index lib/copyscape.rb:52
+     6.0: SI::CopyScape#initialize         lib/copyscape.rb:11
+     5.0: SI::CopyScape#none
+
+    25.8: SI::CopyscapeAPI total
+    14.8: SI::CopyscapeAPI#request         lib/copyscape_api.rb:12
+    11.0: SI::CopyscapeAPI#_call_api       lib/copyscape_api.rb:27
+
+     7.1: SI::CopyscapeResponse total
+     7.1: SI::CopyscapeResponse#results    lib/copyscape_response.rb:27
+```

data/lib/copyscape.rb

@@ -0,0 +1,70 @@
+module SI
+  class CopyScape
+
+    # Data Structures used for responses to the public interface
+    Match = Struct.new(:words_matched, :percent_matched, :title, :url, :copyscape_url, :text_snippet, :html_snippet)
+    Balance = Struct.new(:value, :total, :today)
+    PrivateIndex = Struct.new(:words, :handle, :id, :title)
+
+    attr_reader :api, :match_percent
+
+    def initialize username: nil, api_key: nil, uri: nil, match_percent: 5
+      username ||= ENV['COPYSCAPE_USERNAME']
+      api_key ||= ENV['COPYSCAPE_API_KEY']
+      uri ||= 'http://www.copyscape.com/api/'
+      @match_percent = match_percent
+      @api = SI::CopyscapeAPI.new(username: username, api_key: api_key, api_url: uri)
+    end
+
+    def error
+      api.response.error
+    end
+
+    def credit_balance
+      res = _request(operation: 'balance').remaining
+      Balance.new(res['value'].to_f, res['total'].to_i, res['today'].to_i)
+    end
+
+    def internet_matches! text
+      # cost 1 credit
+      return SI::CopyscapeMatches.new(
+        response: _text_search(text: text, operation: 'csearch'),
+        match_percent: match_percent
+      )
+    end
+
+    def private_matches! text
+      # cost 1 credit
+      return SI::CopyscapeMatches.new(
+        response: _text_search(text: text, operation: 'psearch'),
+        match_percent: match_percent
+      )
+    end
+
+    def internet_and_private_matches! text
+      # cost 2 credits
+      return SI::CopyscapeMatches.new(
+        response: _text_search(text: text, operation: 'cpsearch'),
+        match_percent: match_percent
+      )
+    end
+
+    def add_to_private_index text:, title: nil, id: nil, encoding: 'UTF-8'
+      params = { e: encoding, a: title, i: id }
+      res = _request(operation: 'pindexadd', params: params, postdata: text).response
+      PrivateIndex.new(res['words'].to_i, res['handle'], res['id'], res['title'])
+    end
+
+  private
+
+    def _text_search text:, operation:, encoding: 'UTF-8', full: 3
+      params = { e: encoding, c: full.to_s }
+      _request(operation: operation, params: params, postdata: text);
+    end
+
+    def _request operation:, params: {}, postdata: nil
+      api.request(operation: operation, params: params, postdata: postdata)
+    end
+
+  end
+end

data/lib/copyscape_api.rb

@@ -0,0 +1,39 @@
+module SI
+  class CopyscapeAPI
+
+    attr_reader :api_url, :api_key, :username, :response
+
+    def initialize username:, api_key:, api_url:
+      @username ||= username
+      @api_key ||= api_key
+      @api_url ||= api_url
+    end
+
+    def request operation:, params: {}, postdata: nil
+      uri_hash = { u: username, k: api_key, o: operation }
+      uri_string = api_url + '?' + params.merge(uri_hash).map{|k,v| "#{k}=#{v}"}.join('&')
+      uri = URI.parse(URI.encode(uri_string))
+      @response = _respond_to _call_api(uri, postdata)
+    end
+
+  private
+
+    def _respond_to request
+      if request.is_a?(Net::HTTPSuccess) || request.is_a(Net::HTTPRedirection)
+        SI::CopyscapeResponse.new(raw_response: request)
+      end
+    end
+
+    def _call_api uri, postdata=nil
+      if postdata.nil?
+          Net::HTTP.get_response(uri)
+        else
+          http = Net::HTTP.new(uri.host, uri.port)
+          request = Net::HTTP::Post.new(uri.request_uri)
+          request.body = postdata
+          http.request(request)
+        end
+    end
+
+  end
+end

data/lib/copyscape_matches.rb

@@ -0,0 +1,52 @@
+module SI
+  class CopyscapeMatches
+
+    include Enumerable
+    attr_reader :collection, :query_word_count, :match_percent, :error
+
+    def initialize response:, match_percent:
+      @query_word_count = response.query_words.to_i
+      @match_percent = match_percent.to_i
+      @collection = _build_collection response.results
+      @error = response.error
+    end
+
+    def all_text_snippets
+      collection.map{|m| m.text_snippet}
+    end
+
+    def all_html_snippets
+      collection.map{|m| m.html_snippet}
+    end
+
+    def each(&block)
+      collection.each(&block)
+    end
+
+  private
+
+    def _build_collection results
+      results = _without_rejects(results).map do |match|
+        SI::CopyScape::Match.new(
+          match['wordsmatched'].to_i,
+          match['percentmatched'].to_i,
+          match['title'],
+          match['url'],
+          match['viewurl'],
+          match['textsnippet'],
+          match['htmlsnippet']
+        )
+      end
+      _sort results
+    end
+
+    def _sort results
+      results.sort{|a,b| b.percent_matched <=> a.percent_matched }
+    end
+
+    def _without_rejects results
+      results.reject{|r| r['percentmatched'].to_i < match_percent || !r['urlerror'].nil? }
+    end
+
+  end
+end

data/lib/copyscape_response.rb

@@ -0,0 +1,47 @@
+module SI
+  class CopyscapeResponse
+
+    attr_reader :raw_response, :error
+
+    def initialize raw_response:
+      @raw_response ||= raw_response
+      @error = _error_msg
+    end
+
+    def raw_xml
+      raw_response.body
+    end
+
+    def raw_hash
+      _to_hash
+    end
+
+    def remaining
+      raw_hash['remaining']
+    end
+
+    def response
+      raw_hash['response']
+    end
+
+    def results
+      result = response['result'] if response.is_a?(Hash)
+      result.is_a?(Array) ? result : [result].compact
+    end
+
+    def query_words
+      response['querywords'].to_i if response.is_a?(Hash)
+    end
+
+  private
+
+    def _to_hash
+      Crack::XML.parse(raw_xml)
+    end
+
+    def _error_msg
+      response['error'] if response.is_a?(Hash)
+    end
+
+  end
+end