textrazor 0.0.7 → 0.0.8

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 17c7396459d35be185762cdd33e48422c550382f
-  data.tar.gz: 5162e912ea4c29574f42a291c08945d4be062f03
+  metadata.gz: 5a18bfa0d8979cd6af68c481ff1200802d989dfc
+  data.tar.gz: 71666356571e6e4d6cad269ad6f7f6da306d90d0
 SHA512:
-  metadata.gz: e77134b8e36ed76bd5c779100cb203b2542cebd86d436d7054bb64b695f06aa1509077d811b838246cf86531cd90134562e9d6f225cf83b5e7e7d9b8ab797c50
-  data.tar.gz: 18b6342068846a68bd645599fb2a6c55da017320a1fe2d11cf7386ea192d4998431368cdd3acf72d360632fe9518dfe3f1b5e38af7baf774975b2df3b9c626f4
+  metadata.gz: 653c8e38cf0d1ab4b42796eec8853bc1104c5122ba6e87e6073f74fb2e75b76b75911136a2583f0f0320ab93711da5d269a6ee67280a6c90b3884ca1b7c24961
+  data.tar.gz: c65e0182196f36cf12a080772dd4953ec8d239a6bc66cd3897ef97e2bb83c8bdea5823ef14c91e76d0249980628ca762ae734f1d445c3238f6a5de86c88bc990

data/README.md

@@ -48,6 +48,8 @@ response.entities # Returns an array of TextRazor::Entity instances.
 
 response.words # Returns an array of TextRazor::Word instances.
 
+response.phrases # Returns an array of TextRazor::Phrase instances.
+
 ```
 
 ### One off requests
@@ -65,11 +67,13 @@ TextRazor.entities('api_key', 'text')
 
 TextRazor.words('api_key', 'text')
 
+TextRazor.phrases('api_key', 'text')
+
 ```
 
 ## Next steps
 
-Only implemented this for topics, entities, and words. Also, implement
+Only implemented this for topics, entities, words and phrases. Also, implement
 it for other information that we can retrieve from the public API.
 
 

data/Rakefile

@@ -1,4 +1,5 @@
 require 'bundler/gem_tasks'
+require 'bundler/setup'
 require 'rspec/core/rake_task'
 
 RSpec::Core::RakeTask.new(:spec)

data/lib/textrazor.rb

@@ -1,11 +1,13 @@
 require "textrazor/version"
 require "textrazor/configuration"
+require "textrazor/util"
 require "textrazor/client"
 require "textrazor/request"
 require "textrazor/response"
 require "textrazor/topic"
 require "textrazor/entity"
 require "textrazor/word"
+require "textrazor/phrase"
 
 module TextRazor
 
@@ -21,6 +23,10 @@ module TextRazor
     Client.words(api_key, text, options)
   end
 
+  def self.phrases(api_key, text, options = {})
+    Client.phrases(api_key, text, options)
+  end
+
   class << self
     attr_writer :configuration
   end

data/lib/textrazor/client.rb

@@ -9,7 +9,8 @@ module TextRazor
     DEFAULT_EXTRACTORS = ['entities', 'topics', 'words', 'phrases', 'dependency-trees',
                           'relations', 'entailments', 'senses']
 
-    REQUEST_OPTIONS = [:extractors, :cleanup_html, :language, :filter_dbpedia_types, :filter_freebase_types]
+    REQUEST_OPTIONS = [:extractors, :cleanup_html, :language, 
+                       :filter_dbpedia_types, :filter_freebase_types]
 
     attr_reader :response, :api_key, :request_options
 
@@ -49,6 +50,12 @@ module TextRazor
         entities
     end
 
+    def self.phrases(api_key, text, options = {})
+      new(api_key, options.merge(extractors: ['phrases', 'words'])).
+        analyse(text).
+        phrases
+    end
+
     private
 
     def assign_api_key(api_key)

data/lib/textrazor/entity.rb

@@ -2,6 +2,8 @@ module TextRazor
 
   class Entity
 
+    extend Util
+
     attr_reader :id, :type, :matching_tokens, :entity_id, :freebase_types, :confidence_score,
       :wiki_link, :matched_text, :freebase_id, :relevance_score, :entity_english_id,
       :starting_pos, :ending_pos
@@ -14,19 +16,10 @@ module TextRazor
     end
 
     def self.create_from_hash(params)
-      params = Hash[params.map {|k, v| [underscore(k), v] }]
+      params = Hash[params.map {|k, v| [standardize(k), v] }]
       new(params)
     end
 
-    def self.underscore(text)
-      text.gsub(/::/, '/').
-      gsub(/([A-Z]+)([A-Z][a-z])/,'\1_\2').
-      gsub(/([a-z\d])([A-Z])/,'\1_\2').
-      tr("-", "_").
-      downcase
-    end
-    private_class_method :underscore
-
   end
 
 end

data/lib/textrazor/phrase.rb

@@ -0,0 +1,24 @@
+module TextRazor
+
+  class Phrase
+
+    attr_reader :id, :text
+
+    def initialize(params, words)
+      @id    = params["id"]
+      @text  = match_words(params["wordPositions"], words)
+    end
+
+    def self.create_from_hash(params, words)
+      new(params, words)
+    end
+
+    def match_words(positions, words)
+      phrase = []
+      positions.each { |position| phrase.push words[position].token }
+      phrase.join(' ')
+    end
+
+  end
+
+end

data/lib/textrazor/response.rb

@@ -55,6 +55,10 @@ module TextRazor
       end
     end
 
+    def phrases
+      @phrases ||= parse_phrases(raw_response["nounPhrases"], words)
+    end
+
     private
 
     def bad_request?(code)
@@ -77,6 +81,14 @@ module TextRazor
       end
     end
 
+    def parse_phrases(raw_phrases, words)
+      return nil if raw_phrases.nil?
+
+      raw_phrases.map do |phrase_hash|
+        Phrase.create_from_hash(phrase_hash, words)
+      end
+    end
+
   end
 
 end

data/lib/textrazor/util.rb

@@ -0,0 +1,11 @@
+module TextRazor
+  module Util
+    def standardize(param)
+      param.gsub(/::/, '/').
+      gsub(/([A-Z]+)([A-Z][a-z])/,'\1_\2').
+      gsub(/([a-z\d])([A-Z])/,'\1_\2').
+      tr("-", "_").
+      downcase
+    end
+  end
+end

data/lib/textrazor/version.rb

@@ -1,3 +1,3 @@
 module TextRazor
-  VERSION = "0.0.7"
+  VERSION = "0.0.8"
 end

data/lib/textrazor/word.rb

@@ -2,20 +2,20 @@ module TextRazor
 
   class Word
 
-    attr_reader :position, :starting_pos, :ending_pos, :stem,
-                :lemma, :token, :part_of_speech
+    extend Util
 
-    def initialize(params)
-      @position = params["position"]
-      @starting_pos = params["startingPos"]
-      @ending_pos = params["endingPos"]
-      @stem = params["stem"]
-      @lemma = params["lemma"]
-      @token = params["token"]
-      @part_of_speech = params["partOfSpeech"]
+    attr_reader :position, :starting_pos, :ending_pos, :stem, :lemma, 
+                :token, :part_of_speech, :parent_position
+
+    def initialize(params = {})
+      @type = []
+      params.each do |k, v|
+        instance_variable_set(:"@#{k}", v) if v && self.respond_to?(:"#{k}")
+      end
     end
 
     def self.create_from_hash(params)
+      params = Hash[params.map {|k, v| [standardize(k), v] }]
       new(params)
     end
 

data/spec/lib/textrazor/client_spec.rb

@@ -70,12 +70,12 @@ module TextRazor
         it "should make correct calls" do
           request = Object.new
 
-          Request.should_receive(:post).
+          expect(Request).to receive(:post).
             with('text', {api_key: 'api_key', extractors: %w(entities topics words), cleanup_html: true,
                           language: 'fre', filter_dbpedia_types: %w(type1), filter_freebase_types: %w(type2)}).
             and_return(request)
 
-          Response.should_receive(:new).with(request)
+          expect(Response).to receive(:new).with(request)
 
           client.analyse('text')
         end
@@ -125,11 +125,11 @@ module TextRazor
         client = OpenStruct.new
         response = OpenStruct.new topics: ['topic1'], coarseTopics: ['topic1']
 
-        Client.should_receive(:new).
+        expect(Client).to receive(:new).
           with(api_key, {extractors: ['topics']}).
           and_return(client)
 
-        client.should_receive(:analyse).
+        expect(client).to receive(:analyse).
           with("text").
           and_return(response)
 
@@ -144,11 +144,11 @@ module TextRazor
         client = OpenStruct.new
         response = OpenStruct.new topics: ['topic1'], coarseTopics: ['topic1']
 
-        Client.should_receive(:new).
+        expect(Client).to receive(:new).
           with(api_key, {extractors: ['topics']}).
           and_return(client)
 
-        client.should_receive(:analyse).
+        expect(client).to receive(:analyse).
           with("text").
           and_return(response)
 
@@ -163,11 +163,11 @@ module TextRazor
         client = OpenStruct.new
         response = OpenStruct.new entities: ['Entity1']
 
-        Client.should_receive(:new).
+        expect(Client).to receive(:new).
           with(api_key, {extractors: ['entities']}).
           and_return(client)
 
-        client.should_receive(:analyse).
+        expect(client).to receive(:analyse).
           with("text").
           and_return(response)
 
@@ -182,11 +182,11 @@ module TextRazor
         client = OpenStruct.new
         response = OpenStruct.new words: ['Word1']
 
-        Client.should_receive(:new).
+        expect(Client).to receive(:new).
           with(api_key, {extractors: ['words']}).
           and_return(client)
 
-        client.should_receive(:analyse).
+        expect(client).to receive(:analyse).
           with("text").
           and_return(response)
 
@@ -195,6 +195,25 @@ module TextRazor
 
     end
 
+    context ".phrases" do
+
+      it "should make correct calls" do
+        client = OpenStruct.new
+        response = OpenStruct.new phrases: ['Phrase1']
+
+        expect(Client).to receive(:new).
+          with(api_key, {extractors: ['phrases', 'words']}).
+          and_return(client)
+
+        expect(client).to receive(:analyse).
+          with("text").
+          and_return(response)
+
+        Client.phrases(api_key, 'text', {})
+      end
+
+    end
+
   end
 
 end

data/spec/lib/textrazor/phrase_spec.rb

@@ -0,0 +1,26 @@
+require "spec_helper"
+
+module TextRazor
+
+  describe Phrase do
+
+    context "#create_from_hash" do
+
+      it "should create a new instance" do
+        phrase_hash = {"id" => 1, "wordPositions" => [0, 1, 2]}
+        words       = [
+          OpenStruct.new(token: 'Word1'),
+          OpenStruct.new(token: 'Word2'),
+          OpenStruct.new(token: 'Word3')
+        ]
+
+        phrase = Phrase.create_from_hash(phrase_hash, words)
+
+        expect(phrase.id).to eq(1)
+      end
+
+    end
+
+  end
+
+end

data/spec/lib/textrazor/request_spec.rb

@@ -45,7 +45,7 @@ module TextRazor
         it "should make correct calls" do
           options = {api_key: 'api_key', extractors: %w(entities topics words dependency-trees relations entailments)}
 
-          ::RestClient.should_receive(:post).
+          expect(::RestClient).to receive(:post).
             with("https://api.textrazor.com/", { "text" => 'text', "apiKey" => 'api_key',
             "extractors" => "entities,topics,words,dependency-trees,relations,entailments" }, accept_encoding: 'gzip')
 
@@ -60,7 +60,7 @@ module TextRazor
           options = {api_key: 'api_key', extractors: %w(entities topics words), cleanup_html: true,
                      language: 'fre', filter_dbpedia_types: %w(type1), filter_freebase_types: %w(type2)}
 
-          ::RestClient.should_receive(:post).
+          expect(::RestClient).to receive(:post).
             with("https://api.textrazor.com/", { "text" => 'text', "apiKey" => 'api_key', "extractors" => "entities,topics,words",
             "cleanupHTML" => true, "languageOverride" => 'fre', "entities.filterDbpediaTypes" => "type1",
             "entities.filterFreebaseTypes" => "type2" },

data/spec/lib/textrazor/response_spec.rb

@@ -12,7 +12,7 @@ module TextRazor
           body = "{\"response\":\"{}\"}"
           http_response = ::OpenStruct.new code: 200, body: body
 
-          JSON.should_receive(:parse).
+          expect(JSON).to receive(:parse).
             with(body).
             and_return({"response"=>"{}"})
 

data/spec/lib/textrazor/word_spec.rb

@@ -7,8 +7,16 @@ module TextRazor
     context "#create_from_hash" do
 
       it "should create a new instance" do
-        word_hash = {"position" => 0, "startingPos" => 0, "endingPos" => 3, "stem" => "the",
-                     "lemma" => "the", "token" => "The", "partOfSpeech" => "DT"}
+        word_hash = {
+          "position"        => 0,
+          "startingPos"     => 0,
+          "endingPos"       => 3,
+          "stem"            => "the",
+          "lemma"           => "the",
+          "token"           => "The",
+          "partOfSpeech"    => "DT",
+          "parentPosition"  => 2
+        }
 
         word = Word.create_from_hash(word_hash)
 
@@ -19,6 +27,7 @@ module TextRazor
         expect(word.lemma).to eq("the")
         expect(word.token).to eq("The")
         expect(word.part_of_speech).to eq("DT")
+        expect(word.parent_position).to eq(2)
       end
 
     end

data/spec/lib/textrazor_spec.rb

@@ -5,7 +5,7 @@ describe TextRazor do
   describe ".topics" do
 
     it "should make correct calls" do
-      TextRazor::Client.should_receive(:topics).
+      expect(TextRazor::Client).to receive(:topics).
         with('api_key', 'text', {})
 
       TextRazor.topics('api_key', 'text', {})
@@ -16,7 +16,7 @@ describe TextRazor do
   describe ".entities" do
 
     it "should make correct calls" do
-      TextRazor::Client.should_receive(:entities).
+      expect(TextRazor::Client).to receive(:entities).
         with('api_key', 'text', {})
 
       TextRazor.entities('api_key', 'text', {})
@@ -27,7 +27,7 @@ describe TextRazor do
   describe ".words" do
 
     it "should make correct calls" do
-      TextRazor::Client.should_receive(:words).
+      expect(TextRazor::Client).to receive(:words).
         with('api_key', 'text', {})
 
       TextRazor.words('api_key', 'text', {})
@@ -35,6 +35,17 @@ describe TextRazor do
 
   end
 
+  describe ".phrases" do
+
+    it "should make correct calls" do
+      expect(TextRazor::Client).to receive(:phrases).
+        with('api_key', 'text', {})
+
+      TextRazor.phrases('api_key', 'text', {})
+    end
+
+  end
+
   describe ".reset" do
     before :each do
       TextRazor.configure do |config|

data/spec/spec_helper.rb

@@ -2,9 +2,4 @@ require "bundler"
 Bundler.require
 
 require 'ostruct'
-require 'rspec/fire'
 require File.expand_path("../../lib/textrazor" ,__FILE__)
-
-RSpec.configure do |config|
-  config.include(RSpec::Fire)
-end

data/textrazor.gemspec

@@ -22,7 +22,6 @@ Gem::Specification.new do |spec|
   spec.add_dependency "fast_open_struct"
 
   spec.add_development_dependency "bundler", "~> 1.3"
-  spec.add_development_dependency "rake"
-  spec.add_development_dependency "rspec"
-  spec.add_development_dependency "rspec-fire"
+  spec.add_development_dependency "rake", "~> 10.4"
+  spec.add_development_dependency "rspec", "~> 3.0.0"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: textrazor
 version: !ruby/object:Gem::Version
-  version: 0.0.7
+  version: 0.0.8
 platform: ruby
 authors:
 - Anuj Dutta
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2014-08-06 00:00:00.000000000 Z
+date: 2015-04-04 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rest-client
@@ -56,44 +56,30 @@ dependencies:
   name: rake
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - ">="
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: '0'
+        version: '10.4'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - ">="
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: '0'
+        version: '10.4'
 - !ruby/object:Gem::Dependency
   name: rspec
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - ">="
-      - !ruby/object:Gem::Version
-        version: '0'
-  type: :development
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
-    requirements:
-    - - ">="
-      - !ruby/object:Gem::Version
-        version: '0'
-- !ruby/object:Gem::Dependency
-  name: rspec-fire
-  requirement: !ruby/object:Gem::Requirement
-    requirements:
-    - - ">="
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: '0'
+        version: 3.0.0
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - ">="
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: '0'
+        version: 3.0.0
 description: Api wrapper for text razor
 email:
 - anuj@andhapp.com
@@ -111,14 +97,17 @@ files:
 - lib/textrazor/client.rb
 - lib/textrazor/configuration.rb
 - lib/textrazor/entity.rb
+- lib/textrazor/phrase.rb
 - lib/textrazor/request.rb
 - lib/textrazor/response.rb
 - lib/textrazor/topic.rb
+- lib/textrazor/util.rb
 - lib/textrazor/version.rb
 - lib/textrazor/word.rb
 - spec/lib/textrazor/client_spec.rb
 - spec/lib/textrazor/configuration_spec.rb
 - spec/lib/textrazor/entity_spec.rb
+- spec/lib/textrazor/phrase_spec.rb
 - spec/lib/textrazor/request_spec.rb
 - spec/lib/textrazor/response_spec.rb
 - spec/lib/textrazor/topic_spec.rb
@@ -146,7 +135,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
       version: '0'
 requirements: []
 rubyforge_project: 
-rubygems_version: 2.2.2
+rubygems_version: 2.4.5
 signing_key: 
 specification_version: 4
 summary: An api wrapper for text razor in ruby
@@ -154,6 +143,7 @@ test_files:
 - spec/lib/textrazor/client_spec.rb
 - spec/lib/textrazor/configuration_spec.rb
 - spec/lib/textrazor/entity_spec.rb
+- spec/lib/textrazor/phrase_spec.rb
 - spec/lib/textrazor/request_spec.rb
 - spec/lib/textrazor/response_spec.rb
 - spec/lib/textrazor/topic_spec.rb