Class: RevisionAnalyzer

Inherits:
Object
  • Object
show all
Defined in:
lib/wikidata/diff/revision_analyzer.rb

Constant Summary collapse

CLAIM_TYPES = 
[:added_claims, :removed_claims, :changed_claims,:added_references, :removed_references, :changed_references,:added_qualifiers, :removed_qualifiers, :changed_qualifiers].freeze
ALIAS_TYPES = 
[:added_aliases, :removed_aliases, :changed_aliases].freeze
LABEL_TYPES = 
[:added_labels, :removed_labels, :changed_labels].freeze
DESCRIPTION_TYPES = 
[:added_descriptions, :removed_descriptions, :changed_descriptions].freeze
[:added_sitelinks, :removed_sitelinks, :changed_sitelinks].freeze
COMMENT_TYPES = 
[:merge_to, :merge_from, :redirect, :undo, :restore, :clear_item].freeze
LEMMA_TYPES = 
[:added_lemmas, :removed_lemmas, :changed_lemmas].freeze
FORM_TYPES = 
[:added_forms, :removed_forms, :changed_forms, :added_representations, :removed_representations, :changed_representations, :added_formclaims, :removed_formclaims, :changed_formclaims].freeze
SENSE_TYPES = 
[:added_senses, :removed_senses, :changed_senses, :added_glosses, :removed_glosses, :changed_glosses, :added_senseclaims, :removed_senseclaims, :changed_senseclaims].freeze
NOT_IN_ITEM = 
[:create_lexeme, :create_property, :added_lemmas, :removed_lemmas, :changed_lemmas, :added_forms, :removed_forms, :changed_forms, :added_senses, :removed_senses, :changed_senses, :added_representations, :removed_representations, :changed_representations, :added_glosses, :removed_glosses, :changed_glosses, :added_formclaims, :removed_formclaims, :changed_formclaims, :added_senseclaims, :removed_senseclaims, :changed_senseclaims].freeze
NOT_IN_PROPERTY = 
[:create_lexeme, :create_item, :added_sitelinks, :removed_sitelinks, :changed_sitelinks, :added_lemmas, :removed_lemmas, :changed_lemmas, :added_forms, :removed_forms, :changed_forms, :added_senses, :removed_senses, :changed_senses, :added_representations, :removed_representations, :changed_representations, :added_glosses, :removed_glosses, :changed_glosses, :added_formclaims, :removed_formclaims, :changed_formclaims, :added_senseclaims, :removed_senseclaims, :changed_senseclaims].freeze
NOT_IN_LEXEME = 
[:create_item, :create_property, :added_sitelinks, :changed_sitelinks, :removed_sitelinks, :added_aliases, :changed_aliases, :removed_aliases, :added_labels, :changed_labels, :removed_labels, :added_descriptions, :changed_descriptions, :removed_descriptions].freeze

Class Method Summary collapse

Class Method Details

.analyze_diff(revision_data) ⇒ Object

This method takes two revisions as input and returns the differences between them.



27
28
29
30
31
32
33
34
35
36
37
38
 
# File 'lib/wikidata/diff/revision_analyzer.rb', line 27

def self.analyze_diff(revision_data)
    model = revision_data[:model]
    diff = {}
    if model == 'wikibase-item'
      item(diff, revision_data)
    elsif model == 'wikibase-property'
      property(diff, revision_data)
    elsif model == 'wikibase-lexeme'
      lexeme(diff, revision_data)
    end
    diff
end

.item(diff, revision_data) ⇒ Object 



40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
 
# File 'lib/wikidata/diff/revision_analyzer.rb', line 40

def self.item(diff, revision_data)
  current_content = revision_data[:current_content]
  parent_content = revision_data[:parent_content]
  comment = revision_data[:comment]


  claim_diff = ClaimAnalyzer.isolate_claims_differences(current_content, parent_content)
  CLAIM_TYPES.each do |change_type|
    diff[change_type] = claim_diff[change_type].length
  end

  alias_diff = AliasAnalyzer.isolate_aliases_differences(current_content, parent_content)
  ALIAS_TYPES.each do |change_type|
    diff[change_type] = alias_diff[change_type].length
  end

  # Calculate label differences
  label_diff = LabelAnalyzer.isolate_labels_differences(current_content, parent_content)
  LABEL_TYPES.each do |change_type|
    diff[change_type] = label_diff[change_type].length
  end

  # Calculate description differences
  description_diff = DescriptionAnalyzer.isolate_descriptions_differences(current_content, parent_content)
  DESCRIPTION_TYPES.each do |change_type|
    diff[change_type] = description_diff[change_type].length
  end

  # Calculate sitelink differences
  sitelink_diff = SitelinkAnalyzer.isolate_sitelinks_differences(current_content, parent_content)
  SITELINK_TYPES.each do |change_type|
    diff[change_type] = sitelink_diff[change_type].length
  end

  phrases = CommentAnalyzer.isolate_comment_differences(comment)
  COMMENT_TYPES.each do |change_type|
    diff[change_type] = phrases[change_type]
  end

  NOT_IN_ITEM.each do |change_type|
    diff[change_type] = 0
  end

  diff[:create_item] = phrases[:create_item]

  diff
end

.lexeme(diff, revision_data) ⇒ Object 



129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
 
# File 'lib/wikidata/diff/revision_analyzer.rb', line 129

def self.lexeme(diff, revision_data)
  current_content = revision_data[:current_content]
  parent_content = revision_data[:parent_content]
  comment = revision_data[:comment]
  
  claim_diff = ClaimAnalyzer.isolate_claims_differences(current_content, parent_content)
  CLAIM_TYPES.each do |change_type|
    diff[change_type] = claim_diff[change_type].length
  end

  forms_diff = FormAnalyzer.isolate_forms_differences(current_content, parent_content)
  FORM_TYPES.each do |change_type|
    diff[change_type] = forms_diff[change_type].length
  end

  # Calculate label differences
  lemmas_diff = LemmaAnalyzer.isolate_lemmas_differences(current_content, parent_content)
  LEMMA_TYPES.each do |change_type|
    diff[change_type] = lemmas_diff[change_type].length
  end

  # Calculate description differences
  senses_diff = SenseAnalyzer.isolate_senses_differences(current_content, parent_content)
  SENSE_TYPES.each do |change_type|
    diff[change_type] = senses_diff[change_type].length
  end

  phrases = CommentAnalyzer.isolate_comment_differences(comment)
  COMMENT_TYPES.each do |change_type|
    diff[change_type] = phrases[change_type]
  end
    
  NOT_IN_LEXEME.each do |change_type|
    diff[change_type] = 0
  end
  diff[:create_lexeme] = phrases[:create_item]
end

.property(diff, revision_data) ⇒ Object 



88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
 
# File 'lib/wikidata/diff/revision_analyzer.rb', line 88

def self.property(diff, revision_data)
  current_content = revision_data[:current_content]
  parent_content = revision_data[:parent_content]
  comment = revision_data[:comment]

  claim_diff = ClaimAnalyzer.isolate_claims_differences(current_content, parent_content)
  CLAIM_TYPES.each do |change_type|
    diff[change_type] = claim_diff[change_type].length
  end

  alias_diff = AliasAnalyzer.isolate_aliases_differences(current_content, parent_content)
  ALIAS_TYPES.each do |change_type|
    diff[change_type] = alias_diff[change_type].length
  end

  # Calculate label differences
  label_diff = LabelAnalyzer.isolate_labels_differences(current_content, parent_content)

  LABEL_TYPES.each do |change_type|
    diff[change_type] = label_diff[change_type].length
  end

  # Calculate description differences
  description_diff = DescriptionAnalyzer.isolate_descriptions_differences(current_content, parent_content)
  DESCRIPTION_TYPES.each do |change_type|
    diff[change_type] = description_diff[change_type].length
  end


  phrases = CommentAnalyzer.isolate_comment_differences(comment)
  COMMENT_TYPES.each do |change_type|
    diff[change_type] = phrases[change_type]
  end

  diff[:create_property] = phrases[:create_item]

  NOT_IN_PROPERTY.each do |change_type|
    diff[change_type] = 0
  end
end