Class: PROIEL::PROIELXML::Validator

Inherits:
Object
  • Object
show all
Defined in:
lib/proiel/proiel_xml/validator.rb

Overview

A validator object that uses an XML schema as well as additional integrity checks to validate a PROIEL XML file. Functionality for loading the XML schema and checking the PROIEL XML version number is found in Schema.

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(filename, aligned_filename = nil) ⇒ Validator

Creates a new validator for a PROIEL XML file.

Parameters:

  • filename (String)

    name of PROIEL XML file to validate

  • aligned_filename (NilClass, String) (defaults to: nil)

    name of PROIEL XML file to validate alignments against


21
22
23
24
25
# File 'lib/proiel/proiel_xml/validator.rb', line 21

def initialize(filename, aligned_filename = nil)
  @filename = filename
  @aligned_filename = aligned_filename
  @errors = []
end

Instance Attribute Details

#errorsObject (readonly)

Returns an array of error messages generated during validation.


14
15
16
# File 'lib/proiel/proiel_xml/validator.rb', line 14

def errors
  @errors
end

Instance Method Details

#has_referential_integrity?true, false

Checks the referential integrity of the PROIEL XML file.

If inconsistencies are found, error messages will be appended to `errors`.

Returns:

  • (true, false)

106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
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
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
# File 'lib/proiel/proiel_xml/validator.rb', line 106

def has_referential_integrity?
  tb = PROIEL::Treebank.new
  tb.load_from_xml(@filename)

  errors = []

  # Pass 1: keep track of all object IDs and look for duplicates
  sentence_ids = {}
  token_ids = {}

  tb.sources.each do |source|
    source.divs.each do |div|
      div.sentences.each do |sentence|
        errors << "Repeated sentence ID #{sentence.id}" if sentence_ids.key?(sentence.id)
        sentence_ids[sentence.id] = true

        sentence.tokens.each do |token|
          errors << "Repeated token ID #{token.id}" if token_ids.key?(token.id)
          token_ids[token.id] = { sentence: sentence.id, div: div.id, source: source.id }
        end
      end
    end
  end

  # Pass 2: check object ID references
  tb.sources.each do |source|
    source.tokens.each do |token|
      # Head IDs and slash IDs should be sentence internal
      check_reference_locality(errors, token, token_ids, :head_id, token.head_id, domain: :sentence, allow_nil: true)

      token.slashes.each do |_, target_id|
        check_reference_locality(errors, token, token_ids, :slash_id, target_id, domain: :sentence, allow_nil: false)
      end

      # Antecedent IDs should be source internal
      check_reference_locality(errors, token, token_ids, :antecedent_id, token.antecedent_id, domain: :source, allow_nil: true)
    end
  end

  # Pass 3: verify that all features are defined
  # TBD

  # Pass 4: alignment_id on div, sentence or token requires an alignment_id on source
  tb.sources.each do |source|
    if source.alignment_id.nil?
      if source.divs.any?(&:alignment_id) or source.sentences.any?(&:alignment_id) or source.tokens.any?(&:alignment_id)
        errors << "Alignment ID(s) on divs, sentences or tokens without alignment ID on source"
      end
    end
  end

  # Pass 5: if div is aligned, sentences and tokens within should belong
  # to aligned div(s); if sentence aligned, tokens within should belong
  # to aligned sentence(s). Skip if no alignment_id on source (see pass
  # 4) or if aligned source not available.
  if @aligned_filename
    aligned_tb = PROIEL::Treebank.new
    aligned_tb.load_from_xml(@aligned_filename)

    tb.sources.each do |source|
      if source.alignment_id
        aligned_source = aligned_tb.find_source(source.alignment_id)

        if aligned_source
          check_alignment_integrity(errors, source, aligned_source)
        else
          errors << "Aligned source not available in treebank"
        end
      end
    end
  end

  # Decide if there were any errors
  if errors.empty?
    true
  else
    @errors += errors

    false
  end
end

#valid?true, false

Checks if the PROIEL XML file is valid. This checks for well-formedness, a valid schema version, validation against the schema and referential integrity.

If invalid, `errors` will contain error messages.

Returns:

  • (true, false)

35
36
37
# File 'lib/proiel/proiel_xml/validator.rb', line 35

def valid?
  wellformed? and valid_schema_version? and validates? and has_referential_integrity?
end

#valid_schema_version?true, false

Checks if the PROIEL XML file has a valid schema version number.

If invalid, an error message will be appended to `errors`.

Returns:

  • (true, false)

61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
# File 'lib/proiel/proiel_xml/validator.rb', line 61

def valid_schema_version?
  schema_version = PROIEL::PROIELXML::Schema.check_schema_version_of_xml_file(@filename)

  if schema_version.nil?
    @errors << 'invalid schema version'

    false
  else
    true
  end
rescue PROIEL::PROIELXML::Schema::InvalidSchemaVersion => e
  @errors << e.message

  false
end

#validates?true, false

Checks if the PROIEL XML file validates against the schema.

If invalid, error messages will be appended to `errors`.

Returns:

  • (true, false)

83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
# File 'lib/proiel/proiel_xml/validator.rb', line 83

def validates?
  doc = Nokogiri::XML(File.read(@filename))

  schema_version = PROIEL::PROIELXML::Schema.check_schema_version_of_xml_file(@filename)

  schema = PROIEL::PROIELXML::Schema.load_proiel_xml_schema(schema_version)
  r = schema.validate(doc)

  if r.empty?
    true
  else
    @errors += r.map { |e| "Line #{e.line}: #{e.message}" }

    false
  end
end

#wellformed?true, false

Checks if the PROIEL XML file is well-formed XML.

If not well-formed, an error message will be appended to `errors`.

Returns:

  • (true, false)

45
46
47
48
49
50
51
52
53
# File 'lib/proiel/proiel_xml/validator.rb', line 45

def wellformed?
  Nokogiri::XML(File.read(@filename)) { |config| config.strict }

  true
rescue Nokogiri::XML::SyntaxError => _
  @errors << 'XML file is not wellformed'

  false
end