opener-polarity-tagger 2.4.1 → 3.1.0

data/lib/opener/polarity_tagger/kaf/term.rb

@@ -0,0 +1,35 @@
+module Opener
+  module KAF
+    class Term
+
+      attr_reader :document
+      attr_reader :node
+
+      def initialize document, node
+        @document = document
+        @node     = node
+      end
+
+      def id
+        @id ||= @node.attr :tid
+      end
+
+      def lemma
+        @node.attr :lemma
+      end
+
+      def pos
+        @node.attr :pos
+      end
+
+      def setPolarity attrs, polarity_pos
+        #In case there is no pos info, we use the polarityPos
+        @node[:pos] = polarity_pos if !pos and polarity_pos
+
+        sentiment = @node.add_child('<sentiment/>')
+        sentiment.attr attrs
+      end
+
+    end
+  end
+end

data/lib/opener/polarity_tagger/lexicon_map.rb

@@ -0,0 +1,82 @@
+module Opener
+  class PolarityTagger
+    class LexiconMap
+
+      attr_reader :resource
+      attr_reader :negators
+      attr_reader :intensifiers
+      attr_reader :with_polarity
+
+      POS_ORDER = 'NRVGAO'
+      UNKNOWN   = Hashie::Mash.new polarity: 'unknown'
+
+      def initialize lang:, lexicons:
+        @lang          = lang
+        @lexicons      = lexicons
+
+        @negators      = {}
+        @intensifiers  = {}
+        @with_polarity = {}
+        map lexicons
+      end
+
+      DEFAULT_POS = 'O'
+
+      POS_SHORT_MAP = {
+        adj:         'G',
+        adv:         'A',
+        noun:        'N',
+        propernoun:  'N',
+        other:       'O',
+        prep:        'P',
+        verb:        'V',
+        nil =>       DEFAULT_POS,
+        multi_word_expression: 'O',
+      }
+
+      def by_negator lemma
+        @negators[lemma]
+      end
+
+      def by_intensifier lemma
+        @intensifiers[lemma]
+      end
+
+      def by_polarity lemma, short_pos
+        return [@with_polarity[lemma+short_pos] || UNKNOWN, short_pos] if short_pos
+
+        POS_ORDER.chars.each do |short_pos|
+          if l = @with_polarity[lemma+short_pos]
+            puts "Found polarify #{l.polarity} for #{lemma} with PoS #{short_pos}"
+            return [l, short_pos]
+          end
+        end
+
+        [UNKNOWN, 'unknown']
+      end
+
+      protected
+
+      def map lexicons
+        lexicons.each do |l|
+          next if l.lemma.nil?
+
+          case l.type
+          when 'polarityShifter' then @negators[l.lemma]     = l
+          when 'intensifier'     then @intensifiers[l.lemma] = l
+          else
+            if l.polarity
+              short_pos = POS_SHORT_MAP[l.pos&.to_sym] || DEFAULT_POS
+              @with_polarity[l.lemma+short_pos] = l
+            end
+          end
+        end
+
+        puts "#{@lang}: loaded #{@negators.size} negators"
+        puts "#{@lang}: loaded #{@intensifiers.size} intensifiers"
+        puts "#{@lang}: loaded #{@with_polarity.size} elements with polarity"
+      end
+
+    end
+  end
+end

data/lib/opener/polarity_tagger/lexicons_cache.rb

@@ -0,0 +1,67 @@
+module Opener
+  class PolarityTagger
+    class LexiconsCache
+
+      def initialize
+        extend MonitorMixin
+
+        @url   = ENV['POLARITY_LEXICON_URL']
+        @path  = ENV['POLARITY_LEXICON_PATH']
+        @cache = {}
+      end
+
+      def [] lang
+        synchronize do
+          @cache[lang] ||= load_lexicons lang
+        end
+      end
+      alias_method :get, :[]
+
+      def load_lexicons lang
+        lexicons = if @url then load_from_url lang else load_from_path lang end
+
+        LexiconMap.new lang: lang, lexicons: lexicons
+      end
+
+      def load_from_url lang
+        url      = "#{@url}&language_code=#{lang}"
+        puts "#{lang}: loading lexicons from url #{url}"
+        lexicons = JSON.parse HTTPClient.new.get(url).body
+        lexicons = lexicons['data'].map{ |l| Hashie::Mash.new l }
+        lexicons
+      end
+
+      def load_from_path lang
+        @path  ||= 'core/general-lexicons'
+        dir      = "#{@path}/#{lang.upcase}-lexicon"
+        config   = Nokogiri::XML File.read "#{dir}/config.xml"
+        lexicons = []
+
+        config.css(:lexicon).each do |cl|
+          filename = cl.at(:filename).text
+          resource = cl.at(:resource).text
+          xml      = Nokogiri::XML File.read "#{dir}/#{filename}"
+          puts "#{lang}: loading lexicons from the file #{filename}"
+
+          lexicons.concat(xml.css(:LexicalEntry).map do |le|
+            Hashie::Mash.new(
+              resource:   resource,
+              identifier: le.attr(:id),
+              type:       le.attr(:type),
+              lemma:      le.at(:Lemma).attr(:writtenForm).downcase,
+              pos:        le.attr(:partOfSpeech)&.downcase,
+              aspect:     le.at(:Domain)&.attr(:aspect)&.downcase,
+              polarity:   le.at(:Sentiment).attr(:polarity),
+              strength:   le.at(:Sentiment).attr(:strength),
+              confidence_level:   le.at(:Confidence)&.attr(:level),
+              domain_conditional: le.at(:Domain)&.attr(:conditional) == 'yes',
+            )
+          end)
+        end
+
+        lexicons
+      end
+
+    end
+  end
+end

data/lib/opener/polarity_tagger/server.rb

@@ -1,5 +1,3 @@
-require 'sinatra/base'
-require 'httpclient'
 require 'opener/webservice'
 
 module Opener
@@ -7,10 +5,11 @@ module Opener
     ##
     # Polarity tagger server powered by Sinatra.
     #
-    class Server < Webservice
+    class Server < Webservice::Server
       set :views, File.expand_path('../views', __FILE__)
-      text_processor PolarityTagger
-      accepted_params :input
+
+      self.text_processor  = PolarityTagger
+      self.accepted_params = [:input]
     end # Server
   end # PolarityTagger
 end # Opener

data/lib/opener/polarity_tagger/version.rb

@@ -1,5 +1,7 @@
 module Opener
   class PolarityTagger
-    VERSION = '2.4.1'
-  end # PolarityTagger
-end # Opener
+
+    VERSION = '3.1.0'
+
+  end
+end

data/opener-polarity-tagger.gemspec

@@ -9,6 +9,8 @@ Gem::Specification.new do |gem|
   gem.homepage    = 'http://opener-project.github.com/'
   gem.extensions  = ['ext/hack/Rakefile']
 
+  gem.license = 'Apache 2.0'
+
   gem.required_ruby_version = '>= 1.9.2'
 
   gem.files = Dir.glob([
@@ -20,22 +22,24 @@ Gem::Specification.new do |gem|
     '*.gemspec',
     '*_requirements.txt',
     'README.md',
+    'LICENSE.txt',
     'exec/**/*',
     'task/*'
   ]).select { |file| File.file?(file) }
 
   gem.executables = Dir.glob('bin/*').map { |file| File.basename(file) }
 
+  gem.add_dependency 'opener-daemons', '~> 2.2'
+  gem.add_dependency 'opener-webservice', '~> 2.1'
+  gem.add_dependency 'opener-core', '~> 2.2'
+
+  gem.add_dependency 'hashie'
   gem.add_dependency 'rake'
-  gem.add_dependency 'sinatra'
-  gem.add_dependency 'httpclient'
-  gem.add_dependency 'puma'
-  gem.add_dependency 'opener-daemons'
-  gem.add_dependency 'opener-webservice'
-  gem.add_dependency 'opener-core', '~> 1.0.2'
   gem.add_dependency 'nokogiri'
   gem.add_dependency 'cliver'
+  gem.add_dependency 'slop', '~> 3.5'
 
+  gem.add_development_dependency 'pry'
   gem.add_development_dependency 'rspec', '~> 3.0'
   gem.add_development_dependency 'cucumber'
 end

data/task/requirements.rake

@@ -3,5 +3,5 @@ task :requirements do
   require 'cliver'
 
   Cliver.detect!('python', '~> 2.6')
-  Cliver.detect!('pip', '~> 1.3')
+  Cliver.detect!('pip', '>= 1.3')
 end

metadata

@@ -1,180 +1,183 @@
 --- !ruby/object:Gem::Specification
 name: opener-polarity-tagger
 version: !ruby/object:Gem::Version
-  version: 2.4.1
+  version: 3.1.0
 platform: ruby
 authors:
 - development@olery.com
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2014-06-19 00:00:00.000000000 Z
+date: 2020-09-22 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
-  name: rake
+  name: opener-daemons
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '2.2'
+  type: :runtime
+  prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: '0'
+        version: '2.2'
+- !ruby/object:Gem::Dependency
+  name: opener-webservice
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: '0'
-  prerelease: false
+        version: '2.1'
   type: :runtime
-- !ruby/object:Gem::Dependency
-  name: sinatra
+  prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: '0'
+        version: '2.1'
+- !ruby/object:Gem::Dependency
+  name: opener-core
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: '0'
-  prerelease: false
+        version: '2.2'
   type: :runtime
-- !ruby/object:Gem::Dependency
-  name: httpclient
+  prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: '0'
+        version: '2.2'
+- !ruby/object:Gem::Dependency
+  name: hashie
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '0'
-  prerelease: false
   type: :runtime
-- !ruby/object:Gem::Dependency
-  name: puma
+  prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '0'
+- !ruby/object:Gem::Dependency
+  name: rake
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '0'
-  prerelease: false
   type: :runtime
-- !ruby/object:Gem::Dependency
-  name: opener-daemons
+  prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '0'
+- !ruby/object:Gem::Dependency
+  name: nokogiri
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '0'
-  prerelease: false
   type: :runtime
-- !ruby/object:Gem::Dependency
-  name: opener-webservice
+  prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '0'
+- !ruby/object:Gem::Dependency
+  name: cliver
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '0'
-  prerelease: false
   type: :runtime
-- !ruby/object:Gem::Dependency
-  name: opener-core
+  prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - ~>
+    - - ">="
       - !ruby/object:Gem::Version
-        version: 1.0.2
+        version: '0'
+- !ruby/object:Gem::Dependency
+  name: slop
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - ~>
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: 1.0.2
-  prerelease: false
+        version: '3.5'
   type: :runtime
-- !ruby/object:Gem::Dependency
-  name: nokogiri
+  prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: '0'
+        version: '3.5'
+- !ruby/object:Gem::Dependency
+  name: pry
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '0'
+  type: :development
   prerelease: false
-  type: :runtime
-- !ruby/object:Gem::Dependency
-  name: cliver
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '0'
-  requirement: !ruby/object:Gem::Requirement
-    requirements:
-    - - '>='
-      - !ruby/object:Gem::Version
-        version: '0'
-  prerelease: false
-  type: :runtime
 - !ruby/object:Gem::Dependency
   name: rspec
-  version_requirements: !ruby/object:Gem::Requirement
+  requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - ~>
+    - - "~>"
       - !ruby/object:Gem::Version
         version: '3.0'
-  requirement: !ruby/object:Gem::Requirement
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - ~>
+    - - "~>"
       - !ruby/object:Gem::Version
         version: '3.0'
-  prerelease: false
-  type: :development
 - !ruby/object:Gem::Dependency
   name: cucumber
-  version_requirements: !ruby/object:Gem::Requirement
+  requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '0'
-  requirement: !ruby/object:Gem::Requirement
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '0'
-  prerelease: false
-  type: :development
 description: Polarity tagger for various languages.
 email:
 executables:
 - polarity-tagger
 - polarity-tagger-daemon
 - polarity-tagger-server
+- console
 extensions:
 - ext/hack/Rakefile
 extra_rdoc_files: []
 files:
+- LICENSE.txt
 - README.md
+- bin/console
 - bin/polarity-tagger
 - bin/polarity-tagger-daemon
 - bin/polarity-tagger-server
@@ -197,6 +200,12 @@ files:
 - ext/hack/Rakefile
 - lib/opener/polarity_tagger.rb
 - lib/opener/polarity_tagger/cli.rb
+- lib/opener/polarity_tagger/external.rb
+- lib/opener/polarity_tagger/internal.rb
+- lib/opener/polarity_tagger/kaf/document.rb
+- lib/opener/polarity_tagger/kaf/term.rb
+- lib/opener/polarity_tagger/lexicon_map.rb
+- lib/opener/polarity_tagger/lexicons_cache.rb
 - lib/opener/polarity_tagger/public/markdown.css
 - lib/opener/polarity_tagger/server.rb
 - lib/opener/polarity_tagger/version.rb
@@ -209,7 +218,8 @@ files:
 - task/requirements.rake
 - task/test.rake
 homepage: http://opener-project.github.com/
-licenses: []
+licenses:
+- Apache 2.0
 metadata: {}
 post_install_message:
 rdoc_options: []
@@ -217,17 +227,17 @@ require_paths:
 - lib
 required_ruby_version: !ruby/object:Gem::Requirement
   requirements:
-  - - '>='
+  - - ">="
     - !ruby/object:Gem::Version
       version: 1.9.2
 required_rubygems_version: !ruby/object:Gem::Requirement
   requirements:
-  - - '>='
+  - - ">="
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
 rubyforge_project:
-rubygems_version: 2.2.2
+rubygems_version: 2.7.8
 signing_key:
 specification_version: 4
 summary: Polarity tagger for various languages.