Class: Opener::OpinionDetectorBasic::Kaf::Document

Inherits:
Object
  • Object
show all
Defined in:
lib/opener/opinion_detector_basic/kaf/document.rb

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(file, options = {}) ⇒ Document

Returns a new instance of Document.



8
9
10
11
12
13
14
15
16
 
# File 'lib/opener/opinion_detector_basic/kaf/document.rb', line 8

def initialize file, options = {}
  @document = Nokogiri.XML file

  @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

Dynamic Method Handling

This class handles dynamic methods through the method_missing method

#method_missing(method, *args, &block) ⇒ Object 



139
140
141
 
# File 'lib/opener/opinion_detector_basic/kaf/document.rb', line 139

def method_missing method, *args, &block
  @document.send method, *args, &block
end

Instance Attribute Details

#document ⇒ Object

Returns the value of attribute document.



6
7
8
 
# File 'lib/opener/opinion_detector_basic/kaf/document.rb', line 6

def document
  @document
end

#opinion_strength ⇒ Object

Returns the value of attribute opinion_strength.



6
7
8
 
# File 'lib/opener/opinion_detector_basic/kaf/document.rb', line 6

def opinion_strength
  @opinion_strength
end

#pretty ⇒ Object

Returns the value of attribute pretty.



6
7
8
 
# File 'lib/opener/opinion_detector_basic/kaf/document.rb', line 6

def pretty
  @pretty
end

#timestamp ⇒ Object

Returns the value of attribute timestamp.



6
7
8
 
# File 'lib/opener/opinion_detector_basic/kaf/document.rb', line 6

def timestamp
  @timestamp
end

Instance Method Details

#add_linguistic_processor ⇒ Object

Add linguistic processor layer with basic information (version, timestamp, description etc) in the KAF file.



92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
 
# File 'lib/opener/opinion_detector_basic/kaf/document.rb', line 92

def add_linguistic_processor
  description = 'Basic opinion detector with Pos'
  last_edited = '13may2015'
  version     = '2.0'

  node = new_node('linguisticProcessors', 'KAF/kafHeader')
  node['layer'] = 'opinions'

  lp_node = new_node('lp', node)

  lp_node['version'] = "#{last_edited}-#{version}"
  lp_node['name'] = description

  if timestamp
    format = '%Y-%m-%dT%H:%M:%S%Z'

    lp_node['timestamp'] = Time.now.strftime(format)
  else
    lp_node['timestamp'] = '*'
  end
end

#add_opinion(opinion, index) ⇒ Object

Adds the entire opinion in the KAF file.



38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
 
# File 'lib/opener/opinion_detector_basic/kaf/document.rb', line 38

def add_opinion opinion, index
  opinion_node = new_node 'opinion', 'KAF/opinions'
  opinion_node['oid'] = "o#{index.to_s}"

  if opinion.holders.present?
    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

  if opinion.target_ids.present?
    add_opinion_element opinion_target_node, opinion.target_ids
  end

  expression_node = new_node 'opinion_expression', opinion_node
  expression_node['polarity'] = opinion.polarity
  expression_node['strength'] = opinion.strength.to_s
  expression_node['lexicon-id'] = opinion.lexicon_id if opinion.lexicon_id

  add_opinion_element expression_node, opinion.ids
end

#add_opinion_element(node, ids) ⇒ Object

Method for adding opinion holders, targets and expressions.



75
76
77
78
79
80
81
82
83
84
85
86
 
# File 'lib/opener/opinion_detector_basic/kaf/document.rb', line 75

def add_opinion_element node, ids
  lemmas    = terms.select{|t| ids.include?(t.id)}.map(&:lemma).join(' ')
  comment   = Nokogiri::XML::Comment.new(document, lemmas)
  node.add_child comment

  span_node = new_node('span', node)

  ids.each do |id|
    target_node       = new_node('target', span_node)
    target_node['id'] = id.to_s
  end
end

#add_opinions_layer ⇒ Object

Remove the opinions layer from the KAF file if it exists and add a new one.



64
65
66
67
68
69
70
 
# File 'lib/opener/opinion_detector_basic/kaf/document.rb', line 64

def add_opinions_layer
  existing = document.at_xpath('KAF/opinions')

  existing.remove if existing

  new_node 'opinions', 'KAF'
end

#is_kaf? ⇒ Boolean

Check if input is a KAF file.

Returns:

  • (Boolean) 


135
136
137
 
# File 'lib/opener/opinion_detector_basic/kaf/document.rb', line 135

def is_kaf?
  !!document.at_xpath('KAF')
end

#language ⇒ Object 



24
25
26
 
# File 'lib/opener/opinion_detector_basic/kaf/document.rb', line 24

def language
  @language ||= document.at_xpath('KAF').attr('xml:lang')
end

#new_node(tag, parent) ⇒ Object

Creates a new node in the KAF file.



117
118
119
120
121
122
123
124
125
126
127
128
129
 
# File 'lib/opener/opinion_detector_basic/kaf/document.rb', line 117

def new_node tag, parent
  if parent.is_a?(String)
    parent_node = document.at_xpath(parent)
  else
    parent_node = parent
  end

  node = Nokogiri::XML::Element.new(tag, document)

  parent_node.add_child node

  node
end

#sentences ⇒ Object

Get terms grouped by sentence.



31
32
33
 
# File 'lib/opener/opinion_detector_basic/kaf/document.rb', line 31

def sentences
  @sentences ||= terms.group_by{ |t| t.sentence }
end

#terms ⇒ Object 



18
19
20
21
22
 
# File 'lib/opener/opinion_detector_basic/kaf/document.rb', line 18

def terms
  @terms ||= document.xpath('KAF/terms/term').map do |term|
    Term.new term, self, language
  end
end