opener-opinion-detector-basic 3.0.1 → 3.0.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 3345970ea78e554cf48e14b5293043bc5548d4b3
-  data.tar.gz: 4a1628365168b79760beb9c01d052b81b13c4bc2
+  metadata.gz: fa83717460841775dcf28e9f640208bd16e2cc5d
+  data.tar.gz: 8c8f56b11a38b57c07945bffd666b646ef041f80
 SHA512:
-  metadata.gz: b946c3cce88a034b2ed5952377004f2b5e2122f0d54c9996aa8d71bb6840d752f2e67ce0a059a346c5efcab3072c799a1c09fdc0608b77329fa61df2b3dd1615
-  data.tar.gz: 8dc95b702adee1279a10820f746c99ecd0225308420c48c01ef954bea001650819e05d614bdb20bf4b10b8a13c6f25a839ddca6167e3eae05270eb5e3917476f
+  metadata.gz: 957e43027ccae0ab98becec3641eb82c10a9531d74f278d3557e3f1f8aa19e8ec59260bebd31a105e5e283e55bc2a04d67b70f9a582c00b93b771210e491fbe8
+  data.tar.gz: 02a6df21160370204fb1a72084d36663f974ff816a615028567e701370bd0a6cfabe96b16f7281f0576d7f656bbd21de010f129a05d1f2d3126e71ed4b1706ff

data/exec/opinion-detector-basic.rb

@@ -4,6 +4,44 @@ require 'opener/daemons'
 
 require_relative '../lib/opener/opinion_detector_basic'
 
+Oga::XML::Parser.class_eval do
+  include NewRelic::Agent::Instrumentation::ControllerInstrumentation
+  include NewRelic::Agent::MethodTracer
+
+  add_method_tracer(:parse)
+end
+
+Oga::XPath::Parser.class_eval do
+  class << self
+    include NewRelic::Agent::Instrumentation::ControllerInstrumentation
+    include NewRelic::Agent::MethodTracer
+
+    add_method_tracer(:parse_with_cache, 'Oga::XPath::Parser/parse_with_cache')
+  end
+end
+
+Oga::XPath::Evaluator.class_eval do
+  include NewRelic::Agent::Instrumentation::ControllerInstrumentation
+  include NewRelic::Agent::MethodTracer
+
+  add_method_tracer(:evaluate)
+  add_method_tracer(:evaluate_ast)
+end
+
+Opener::OpinionDetectorBasic::Processor.class_eval do
+  include NewRelic::Agent::Instrumentation::ControllerInstrumentation
+  include NewRelic::Agent::MethodTracer
+
+  add_method_tracer(:process)
+  add_method_tracer(:terms)
+  add_method_tracer(:opinions)
+  add_method_tracer(:add_opinion_element)
+  add_method_tracer(:pretty_print)
+  add_method_tracer(:set_accumulated_strength)
+  add_method_tracer(:apply_modifiers)
+  add_method_tracer(:apply_conjunctions)
+end
+
 daemon = Opener::Daemons::Daemon.new(Opener::OpinionDetectorBasic)
 
 daemon.start

data/lib/opener/opinion_detector_basic.rb

@@ -1,6 +1,7 @@
+gem 'slop', '~> 3.0'
+
 require 'slop'
 require 'oga'
-require 'monitor'
 
 require 'rexml/document'
 require 'rexml/formatters/pretty'
@@ -32,7 +33,7 @@ module Opener
       @args    = options.delete(:args) || []
       @options = options
     end
-    
+
     ##
     # Processes the input KAF document.
     #
@@ -40,8 +41,6 @@ module Opener
     # @return [String]
     #
     def run(input)
-      options[:timestamp] = !options.delete(:no_time)
-
       return Processor.new(input, options).process
     end
   end # OpinionDetectorBasic

data/lib/opener/opinion_detector_basic/cli.rb

@@ -45,10 +45,17 @@ Example:
             abort "opinion-detector-basic v#{VERSION} on #{RUBY_DESCRIPTION}"
           end
 
+          on :p, :pretty, 'Enables pretty-printing of XML',
+            :default => false
+
+          on :'no-timestamp', 'Disables timestamps in XML output'
+
           run do |opts, args|
             detector = OpinionDetectorBasic.new(
-              :args   => args,
-              :domain => opts[:domain]
+              :args      => args,
+              :domain    => opts[:domain],
+              :pretty    => opts[:pretty],
+              :timestamp => !opts[:'no-timestamp']
             )
 
             input = STDIN.tty? ? nil : STDIN.read

data/lib/opener/opinion_detector_basic/opinion.rb

@@ -3,7 +3,22 @@ module Opener
     class Opinion
       attr_reader :term
       attr_accessor :left_candidates, :right_candidates, :target_ids, :holders
-      
+
+      # Opinion holders for each language code.
+      OPINION_HOLDERS = {
+        'nl' => [
+          'ik','we','wij','ze','zij','jullie','u','hij','het','jij','je','mij',
+          'me','hem','haar','ons','hen','hun'
+        ],
+        'en' => ['i','we','he','she','they','it','you'],
+        'es' => [
+          'yo','tu','nosotros','vosotros','ellos','ellas','nosotras','vosotras'
+        ],
+        'it' => ['io','tu','noi','voi','loro','lei','lui'],
+        'de' => ['ich','du','wir','ihr','sie','er'],
+        'fr' => ['je','tu','lui','elle','nous','vous','ils','elles']
+      }
+
       def initialize(term)
         @term = term
         @left_candidates = []
@@ -11,7 +26,7 @@ module Opener
         @holders = []
         @target_ids = []
       end
-      
+
       ##
       # Returns the term ids of the opinion expression.
       #
@@ -20,7 +35,7 @@ module Opener
       def ids
         @ids ||= term.list_ids.sort
       end
-      
+
       ##
       # Returns the sentence id of the opinion.
       #
@@ -29,7 +44,7 @@ module Opener
       def sentence
         @sentence ||= term.sentence
       end
-      
+
       ##
       # Returns the strength of the opinion.
       #
@@ -38,7 +53,7 @@ module Opener
       def strength
         @strength ||= term.accumulated_strength
       end
-      
+
       ##
       # Returns the polarity of the opinion.
       #
@@ -53,7 +68,7 @@ module Opener
           "neutral"
         end
       end
-      
+
       ##
       # Obtain the opinion holders from the terms that belong to the same
       # sentence.
@@ -61,13 +76,13 @@ module Opener
       def obtain_holders(sentences, language)
         sentence_terms = sentences[sentence]
         sentence_terms.each do |term|
-          if opinion_holders[language].include?(term.lemma)
+          if OPINION_HOLDERS[language].include?(term.lemma)
             @holders << term.id
             break
           end
         end
       end
-      
+
       ##
       # Get the potential right and left candidates of the sentence and
       # decide which ones are the actual targets of the opinion
@@ -76,20 +91,20 @@ module Opener
         sentence_terms = sentences[sentence]
         max_distance = 3
         terms_count = sentence_terms.count
-        
+
         index = -1
         sentence_terms.each_with_index do |term, i|
           if ids.include?(term.id)
             index = i
           end
         end
-        
+
         unless index+1 >= terms_count
           min = index+1
           max = [index+1+max_distance,terms_count].min
           @right_candidates = filter_candidates(sentence_terms[min..max])
         end
-        
+
         index = 0
         sentence_terms.each_with_index do |term, i|
           if ids.include?(term.id)
@@ -97,7 +112,7 @@ module Opener
             break # needed for left_candidates
           end
         end
-        
+
         unless index == 0
           min = [0, index-1-max_distance].max
           max = index
@@ -108,7 +123,7 @@ module Opener
           candidate = right_candidates.first
           @target_ids << candidate.id
         end
-        
+
         if target_ids.empty?
           list = mix_lists(right_candidates, left_candidates)
           list.each do |l|
@@ -117,9 +132,9 @@ module Opener
           end
         end
       end
-      
+
       protected
-      
+
       ##
       # If there are no opinion targets, right and left candidates
       # are mixed into one list and the first one is picked as the target.
@@ -140,7 +155,7 @@ module Opener
         end
         return list.compact
       end
-      
+
       ##
       # Filters candidate terms depending on their part of speech and if
       # they are already part of the expression.
@@ -150,22 +165,6 @@ module Opener
       def filter_candidates(sentence_terms)
         sentence_terms.select{|t| (t.pos == "N" || t.pos == "R") && !ids.include?(t.id)}
       end
-      
-      ##
-      # Opinion holders for each language code.
-      #
-      # @return [Hash]
-      #
-      def opinion_holders
-        {
-          'nl' => ['ik','we','wij','ze','zij','jullie','u','hij','het','jij','je','mij','me','hem','haar','ons','hen','hun'],
-          'en' => ['i','we','he','she','they','it','you'],
-          'es' => ['yo','tu','nosotros','vosotros','ellos','ellas','nosotras','vosotras'],
-          'it' => ['io','tu','noi','voi','loro','lei','lui'],
-          'de' => ['ich','du','wir','ihr','sie','er'],
-          'fr' => ['je','tu','lui','elle','nous','vous','ils','elles']
-        }
-      end
     end # Opinion
   end # OpinionDetectorBasic
-end # Opener
+end # Opener

data/lib/opener/opinion_detector_basic/processor.rb

@@ -17,11 +17,11 @@ module Opener
       #  by default due to the performance overhead.
       #
       def initialize(file, options = {})
-        @document            = Oga.parse_xml(file)
+        @document = Oga.parse_xml(file)
 
-        @timestamp           = !!options[:timestamp]
-        @opinion_strength    = !!options[:opinion_strength]
-        @pretty              = options[:pretty] || false
+        @timestamp        = options[:timestamp]
+        @opinion_strength = options[:opinion_strength]
+        @pretty           = options[:pretty] || false
 
         raise 'Error parsing input. Input is required to be KAF' unless is_kaf?
       end
@@ -30,18 +30,18 @@ module Opener
       # Processes the input and returns the new KAF output.
       # @return [String]
       #
-      def process        
+      def process
         add_opinions_layer
-        
+
         index = 1
-        opinions.each do |opinion|          
+        opinions.each do |opinion|
           add_opinion(opinion, index)
           index += 1
         end
 
         add_linguistic_processor
 
-        return pretty ? pretty_print(document) : document.to_xml
+        pretty ? pretty_print(document) : document.to_xml
       end
 
       ##
@@ -50,25 +50,19 @@ module Opener
       # @return [String]
       #
       def language
-        return @language ||= document.at_xpath('KAF').get('xml:lang')
+        @language ||= document.at_xpath('KAF').get('xml:lang')
       end
-      
+
       ##
       # Get the terms from the input file
       # @return [Hash]
       #
       def terms
-        unless @terms
-          @terms = []
-          
-          document.xpath('KAF/terms/term').each do |term|
-            @terms << Term.new(term, document, language)
-          end
+        @terms ||= document.xpath('KAF/terms/term').map do |term|
+          Term.new(term, document, language)
         end
-        
-        return @terms
       end
-      
+
       ##
       # Get the opinions.
       #
@@ -79,7 +73,7 @@ module Opener
           set_accumulated_strength
           apply_modifiers
           apply_conjunctions
-          
+
           ##
           # Initialize opinions with their expressions.
           #
@@ -88,14 +82,14 @@ module Opener
               o = Opinion.new(term)
             end
           end.compact
-          
+
           ##
           # Obtain targets for each opinion.
           #
           @opinions.each do |opinion|
             opinion.obtain_targets(sentences)
           end
-          
+
           ##
           # Obtain holders for each opinion.
           #
@@ -103,10 +97,10 @@ module Opener
             opinion.obtain_holders(sentences, language)
           end
         end
-        
-        return @opinions
+
+        @opinions
       end
-      
+
       ##
       # Remove the opinions layer from the KAF file if it exists and add a new
       # one.
@@ -117,30 +111,32 @@ module Opener
 
         new_node('opinions', 'KAF')
       end
-      
+
       ##
       # Adds the entire opinion in the KAF file.
       #
       def add_opinion(opinion, index)
         opinion_node = new_node("opinion", "KAF/opinions")
         opinion_node.set('oid', "o#{index.to_s}")
-        
+
         unless opinion.holders.empty?
           opinion_holder_node = new_node("opinion_holder", opinion_node)
           add_opinion_element(opinion_holder_node, opinion.holders)
         end
-        
+
         opinion_target_node = new_node("opinion_target", opinion_node)
+
         unless opinion.target_ids.empty?
           add_opinion_element(opinion_target_node, opinion.target_ids)
         end
-        
+
         expression_node = new_node("opinion_expression", opinion_node)
         expression_node.set('polarity', opinion.polarity)
         expression_node.set('strength', opinion.strength.to_s)
+
         add_opinion_element(expression_node, opinion.ids)
       end
-      
+
       ##
       # Method for adding opinion holders, targets and expressions.
       #
@@ -149,12 +145,13 @@ module Opener
         comment = Oga::XML::Comment.new(:text => "#{lemmas}")
         node.children << comment
         span_node = new_node("span", node)
+
         ids.each do |id|
           target_node = new_node("target", span_node)
           target_node.set('id', id.to_s)
         end
       end
-      
+
       ##
       # Add linguistic processor layer with basic information
       # (version, timestamp, description etc) in the KAF file.
@@ -171,7 +168,7 @@ module Opener
 
         lp_node.set('version', "#{last_edited}-#{version}")
         lp_node.set('name', description)
-        
+
         if timestamp
           format = '%Y-%m-%dT%H:%M:%S%Z'
 
@@ -180,7 +177,7 @@ module Opener
           lp_node.set('timestamp', '*')
         end
       end
-      
+
       ##
       # Format the output document properly.
       #
@@ -196,18 +193,18 @@ module Opener
         formatter.compact = true
         formatter.write(doc, out)
 
-        return out.strip
+        out.strip
       end
-      
+
       ##
       # Get terms grouped by sentence.
       #
       def sentences
         @sentences ||= terms.group_by{|t| t.sentence}
       end
-      
+
       protected
-      
+
       ##
       # The strength of a term depends heavily on the type of the previous
       # one. For example if the previous one is a shifter, it needs
@@ -228,7 +225,7 @@ module Opener
                 symbol = terms[i+1].accumulated_strength > 0 ? :+ : :-
               else
                 symbol = :*
-              end          
+              end
             elsif terms[i+1].is_intensifier?
               terms[i+1].accumulated_strength = term.accumulated_strength.send(symbol, terms[i+1].accumulated_strength)
               term.use = false
@@ -242,7 +239,7 @@ module Opener
           end
         end
       end
-      
+
       ##
       # Apply strength to the next term after a shifter or intensifier.
       #
@@ -258,7 +255,7 @@ module Opener
           end
         end
       end
-      
+
       ##
       # Ignore conjunctions when applying strength.
       #
@@ -276,7 +273,7 @@ module Opener
               if j >= terms_count
                 break
               end
-              
+
               if terms[j].is_conjunction
                 terms[j].use = false
                 j += 1
@@ -297,9 +294,9 @@ module Opener
             i = j
           end
           i += 1
-        end    
+        end
       end
-      
+
       ##
       # Creates a new node in the KAF file.
       #
@@ -314,7 +311,7 @@ module Opener
 
         parent_node.children << node
 
-        return node
+        node
       end
 
       ##
@@ -322,7 +319,7 @@ module Opener
       # @return [Boolean]
       #
       def is_kaf?
-        return !!document.at_xpath('KAF')
+        !!document.at_xpath('KAF')
       end
     end # Processor
   end # OpinionDetectorBasic

data/lib/opener/opinion_detector_basic/term.rb

@@ -3,7 +3,17 @@ module Opener
     class Term
       attr_reader :node, :sentence, :is_conjunction
       attr_accessor :use, :accumulated_strength, :list_ids
-      
+
+      # Map of conjunctions per language code
+      CONJUNCTIONS = {
+        'nl' => %w{, en},
+        'en' => %w{, and},
+        'es' => %w{, y e},
+        'it' => %w{, e ed},
+        'de' => %w{, und},
+        'fr' => %w{, et}
+      }
+
       def initialize(node, document, language)
         @node                 = node
         @sentence             = get_sentence(document)
@@ -12,7 +22,7 @@ module Opener
         @list_ids             = [id]
         @is_conjunction       = is_conjunction?(language)
       end
-      
+
       ##
       # Returns the term id.
       #
@@ -21,16 +31,16 @@ module Opener
       def id
         @id ||= node.get('tid')
       end
-      
+
       ##
       # Returns the lemma of the term.
-      # 
+      #
       # @return [String]
       #
       def lemma
         @lemma ||= node.get('lemma')
       end
-      
+
       ##
       # Returns the part of speech of the term.
       #
@@ -39,38 +49,36 @@ module Opener
       def pos
         @pos ||= node.get('pos')
       end
-      
+
       ##
       # Returns the sentiment modifier type if it exists.
       #
       # @return [String|NilClass]
       #
       def sentiment_modifier
-        @sentiment_modifier ||= if sentiment = node.xpath('sentiment').first
-          sentiment.get('sentiment_modifier')
-        end
+        @sentiment_modifier ||=
+          first_sentiment ? first_sentiment.get('sentiment_modifier') : nil
       end
-      
+
       ##
       # Returns the polarity of the term if it exists.
       #
       # @return [String|NilClass]
       #
       def polarity
-        @polarity ||= if sentiment = node.xpath('sentiment').first
-          sentiment.get('polarity')
-        end
+        @polarity ||= first_sentiment ? first_sentiment.get('polarity') : nil
       end
-      
+
       ##
       # Returns the actual word ids that construct the lemma.
       #
       # @return [Array]
       #
       def target_ids
-        @target_ids ||= node.xpath('span/target').map {|target| target.get('id')}
+        @target_ids ||= node.xpath('span/target')
+          .map { |target| target.get('id') }
       end
-      
+
       ##
       # Returns the strength of the term depending on its type.
       #
@@ -82,16 +90,16 @@ module Opener
         elsif polarity == "negative"
           return -1
         end
-        
+
         if is_intensifier?
           return 2
         elsif is_shifter?
           return -1
         end
-        
+
         return 0
       end
-      
+
       ##
       # Returns the sentence id that the term belongs to in the document.
       #
@@ -103,7 +111,7 @@ module Opener
         .first
         .get('sent')
       end
-      
+
       ##
       # Checks if a term is an intensifier.
       #
@@ -112,7 +120,7 @@ module Opener
       def is_intensifier?
         sentiment_modifier == "intensifier"
       end
-      
+
       ##
       # Checks if a term is a shifter.
       #
@@ -121,7 +129,7 @@ module Opener
       def is_shifter?
         sentiment_modifier == "shifter"
       end
-      
+
       ##
       # Checks if a term is an expression.
       #
@@ -130,31 +138,22 @@ module Opener
       def is_expression?
         use && !!polarity
       end
-      
+
       ##
       # Checks if a term is a conjunction.
       #
       # @return [TrueClass|FalseClass]
       #
       def is_conjunction?(language)
-        conjunctions[language].include?(lemma)
+        CONJUNCTIONS[language].include?(lemma)
       end
-      
-      ##
-      # Map of conjunctions per language code
-      #
-      # @return [Hash]
-      #
-      def conjunctions
-        {
-          'nl' => [',','en'],
-          'en' => [',','and'],
-          'es' => [',','y','e'],
-          'it' => [',','e','ed'],
-          'de' => [',','und'],
-          'fr' => [',','et']         
-        }
+
+      private
+
+      # @return [Oga::XML::Element]
+      def first_sentiment
+        @first_sentiment ||= node.xpath('sentiment').first
       end
     end # Term
   end # OpinionDetectorBasic
-end # Opener
+end # Opener

data/lib/opener/opinion_detector_basic/version.rb

@@ -1,5 +1,5 @@
 module Opener
   class OpinionDetectorBasic
-    VERSION = '3.0.1'
+    VERSION = '3.0.2'
   end
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: opener-opinion-detector-basic
 version: !ruby/object:Gem::Version
-  version: 3.0.1
+  version: 3.0.2
 platform: ruby
 authors:
 - development@olery.com
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2015-06-15 00:00:00.000000000 Z
+date: 2015-08-18 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: opener-daemons
@@ -125,9 +125,9 @@ dependencies:
 description: Basic Opinion Detector.
 email: 
 executables:
+- opinion-detector-basic-server
 - opinion-detector-basic
 - opinion-detector-basic-daemon
-- opinion-detector-basic-server
 extensions: []
 extra_rdoc_files: []
 files:
@@ -169,7 +169,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
       version: '0'
 requirements: []
 rubyforge_project: 
-rubygems_version: 2.2.2
+rubygems_version: 2.4.8
 signing_key: 
 specification_version: 4
 summary: Basic Opinion Detector.