Class: TAO::RDFizer

Inherits:
Object
  • Object
show all
Defined in:
lib/tao_rdfizer/tao_rdfizer.rb

Instance Method Summary collapse

Constructor Details

#initialize(mode = nil) ⇒ RDFizer

if mode == :spans then produces span descriptions if mode == :annotations then produces annotation descriptions if mode == nil then produces both



10
11
12
13
14
15
16
17
18
19
20
 
# File 'lib/tao_rdfizer/tao_rdfizer.rb', line 10

def initialize(mode = nil)
  @mode = mode
  template = if !mode.nil? && mode == :spans
    ERB_SPANS_TTL
  else
    ERB_ANNOTATIONS_TTL
  end

  @tao_ttl_erb = ERB.new(template, nil, '-')
  @prefix_ttl_erb = ERB.new(ERB_PREFIXES_TTL, nil, '-')
end

Instance Method Details

#rdfize(annotations_col, with_prefix = true) ⇒ Object 



22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
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
87
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
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
 
# File 'lib/tao_rdfizer/tao_rdfizer.rb', line 22

def rdfize(annotations_col, with_prefix = true)
  # check the format
  annotations_col.each do |annotations|
    raise "'target' is missing" unless annotations.has_key? :target
  end

  # namespaces
  namespaces = {}

  anns = annotations_col.first
  anns[:namespaces].each {|n| namespaces[n[:prefix]] = n[:uri]} unless anns[:namespaces].nil?

  unless @mode ==:spans
    raise ArgumentError, "A project name has to be specified." unless anns.has_key?(:project)
    prefix_for_this = anns[:project].downcase.gsub(/ /, '_')
    raise ArgumentError, "'#{prefix_for_this}' is a reserved prefix for this project." if namespaces.has_key?(prefix_for_this)
    project_uri = 'http://pubannotation.org/projects/' + anns[:project]
    namespaces[prefix_for_this] = project_uri + '/'
  end

  denotations = []
  attributes = []
  relations = []
  spans = []

  annotations_col.each do |annotations|
    text = annotations[:text]
    text_uri = annotations[:target]
    text_id = begin
      sourcedb, sourceid, divid = get_target_info(text_uri)
      divid.nil? ? "#{sourcedb}-#{sourceid}" : "#{sourcedb}-#{sourceid}-#{divid}"
    end

    # denotations and relations
    _denotations = annotations[:denotations] || []
    _attributes = annotations[:attributes] || []
    _relations = annotations[:relations] || []
    if @mode == :spans && annotations.has_key?(:tracks)
      annotations[:tracks].each do |track|
        _denotations += track[:denotations]
        _attributes += track[:attributes]
        _relations += track[:relations]
      end
    end

    begin
      if @mode == :annotations
        # index attributes
        attributesh = _attributes.inject({}) do |h, a|
          if a[:pred].end_with?('_id')
            subj = a[:subj]
            h[subj] = [] unless h.has_key? subj
            h[subj] << a[:obj]
          end
          h
        end

        # denotations preprocessing
        _denotations.each do |d|
          span_uri = "<#{text_uri}/spans/#{d[:span][:begin]}-#{d[:span][:end]}>"
          d[:span_uri] = span_uri
          d[:obj_uri] = "#{prefix_for_this}:#{text_id}-#{d[:id]}"
          class_uris = (attributesh[d[:id]] || []).push(d[:obj])
          d[:class_uris] = class_uris.map{|uri| find_uri(uri, namespaces, prefix_for_this)}
        rescue ArgumentError => e
          raise ArgumentError, "[#{sourcedb}-#{sourceid}-#{d[:id]}] " + e.message
        end

        # relations preprocessing
        _relations.each do |r|
          r[:subj_uri] = "#{prefix_for_this}:#{text_id}-#{r[:subj]}"
          r[:obj_uri] = "#{prefix_for_this}:#{text_id}-#{r[:obj]}"
          r[:pred_uri] = find_uri(r[:pred], namespaces, prefix_for_this)
        rescue ArgumentError => e
          raise ArgumentError, "[#{sourcedb}-#{sourceid}-#{r[:id]}] " + e.message
        end
      else
        # denotations preprocessing
        _denotations.each do |d|
          span_uri = "<#{text_uri}/spans/#{d[:span][:begin]}-#{d[:span][:end]}>"
          d[:span_uri] = span_uri
        rescue ArgumentError => e
          raise ArgumentError, "[#{sourcedb}-#{sourceid}-#{d[:id]}] " + e.message
        end
      end
    end

    unless @mode == :annotations
      # collect spans
      _spans = _denotations.map{|d| d[:span]}
      _spans.uniq!

      # add_infomation
      _spans.each do |s|
        s[:span_uri] = "<#{text_uri}/spans/#{s[:begin]}-#{s[:end]}>"
        s[:source_uri] = text_uri
        s[:text] = text[s[:begin] ... s[:end]]
      end

      # index spans
      spanh = _spans.inject({}){|r, s| r[s[:span_uri]] = s; r}

      # add denotation information
      _denotations.each do |d|
        span_uri = d[:span_uri]
        if spanh[span_uri][:denotations].nil?
          spanh[span_uri][:denotations] = [d]
        else
          spanh[span_uri][:denotations] << d
        end
      end

      _spans.sort!{|a, b| (a[:begin] <=> b[:begin]).nonzero? || b[:end] <=> a[:end]}

      ## begin indexing
      len = text.length
      num = _spans.length

      # initilaize the index
      (0 ... num).each do |i|
        _spans[i][:followings] = []
        _spans[i][:children] = []
      end

      (0 ... num).each do |i|
        # find the first following span
        j = i + 1

        while j < num && _spans[j][:begin] < _spans[i][:end]
          unless include_parent?(_spans[i][:children], _spans[j])
            _spans[i][:children] << _spans[j]
          end
          j += 1
        end

        # find adjacent positions
        fp = _spans[i][:end]
        fp += 1 while fp < len && text[fp].match(/\s/)
        next if fp == len

        # index adjacent spans
        while j < num && _spans[j][:begin] == fp
          _spans[i][:followings] << _spans[j]
          j += 1
        end
      end
    end

    denotations += _denotations
    relations += _relations
    spans += _spans unless @mode == :annotations
  end

  ttl  = ''
  ttl += @prefix_ttl_erb.result(binding) if with_prefix
  ttl += @tao_ttl_erb.result(binding)
end