Module: Segment

Extended by:
Annotation
Included in:
NLP::GdepChunk, NLP::GdepToken, NamedEntity, Relationship
Defined in:
lib/rbbt/ner/segment.rb

Defined Under Namespace

Classes: Index

Class Method Summary collapse

Instance Method Summary collapse

Class Method Details

.align(text, parts) ⇒ Object

{{{ Align



172
173
174
175
176
177
178
179
180
181
 
# File 'lib/rbbt/ner/segment.rb', line 172

def self.align(text, parts)
  pre_offset = 0
  parts.each do |part|
    offset = text.index part
    next if offset.nil?
    Segment.setup(part, pre_offset + offset)
    pre_offset += offset + part.segment_length - 1
    text = text[(offset + part.segment_length - 1)..-1]
  end
end

.clean_sort(segments) ⇒ Object 



114
115
116
117
118
119
120
121
122
 
# File 'lib/rbbt/ner/segment.rb', line 114

def self.clean_sort(segments)
  sorted = sort(segments).reject{|s| s.offset.nil?}
  overlaps = overlaps(sorted)
  overlaps.each do |s|
    sorted.delete s
  end

  sorted
end

.index(segments, persist_file = :memory) ⇒ Object 



197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
 
# File 'lib/rbbt/ner/segment.rb', line 197

def self.index(segments, persist_file = :memory)
  segments = segments.values.flatten if Hash === segments

  annotation_index = 
    Persist.persist("Segment_index", :fwt, :persist => (! (persist_file.nil? or persist_file == :memory)), :file => persist_file) do

      value_size = 0
      index_data = segments.collect{|segment| 
        next if segment.offset.nil?
        range = segment.range
        value_size = [segment.id.length, value_size].max
        [segment.id, [range.begin, range.end]]
      }.compact

      fwt = FixWidthTable.get :memory, value_size, true
      fwt.add_range index_data

      fwt
    end

  data = {}
  segments.each do |segment| data[segment.id] = segment end
  Index.new annotation_index, data
end

.load_tsv(tsv) ⇒ Object 



322
323
324
325
326
327
328
329
 
# File 'lib/rbbt/ner/segment.rb', line 322

def self.load_tsv(tsv)
  fields = tsv.fields
  tsv.with_unnamed do
    tsv.collect do |id, values|
      Annotated.load_tsv_values(id, values, fields)
    end
  end
end

.load_tsv_values(text, values, fields) ⇒ Object 



246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
 
# File 'lib/rbbt/ner/segment.rb', line 246

def self.load_tsv_values(text, values, fields)
  info = {}
  literal_pos = fields.index "literal"

  object = if literal_pos.nil?
             ""
           else
             v = values[literal_pos]
             v = v.first if Array === v
             v
           end

  fields.each_with_index do |field, i|
    if field == "JSON"
      JSON.parse(values[i]).each do |key, value|
        info[key.to_sym] = value
      end
    else
      info[field.to_sym] = values[i]
    end
  end

  start = info.delete(:Start)
  if not (start.nil? or ((Array === start or String === start) and start.empty?))
    if Array === start
      start = start.first
    end
    start = start.to_i
    info[:offset] = start

    eend = info.delete(:End)
    if Array === eend
      eend = eend.first
    end
    eend = eend.to_i

    if object.empty?
      object.replace text[start..eend]
    end
  end

  info[:annotation_types] = [Segment] unless info.include? :annotation_types

  Annotated.load(object, info)
end

.overlaps(sorted_segments) ⇒ Object 



102
103
104
105
106
107
108
109
110
111
112
 
# File 'lib/rbbt/ner/segment.rb', line 102

def self.overlaps(sorted_segments)
  last = nil
  overlaped = []

  sorted_segments.reverse.each do |segment| 
    overlaped << segment if (not last.nil?) and segment.range.end > last 
    last = segment.range.begin
  end

  overlaped
end

.set_tsv_fields(fields, segments) ⇒ Object 



292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
 
# File 'lib/rbbt/ner/segment.rb', line 292

def self.set_tsv_fields(fields, segments)
  tsv_fields = []
  add_types = ! (fields.delete(:no_types) || fields.delete("no_types") || fields.include?(:JSON) || fields.include?("JSON"))
  literal = (fields.delete(:literal) || fields.delete("literal"))
  tsv_fields << "Start" << "End"
  tsv_fields << :annotation_types if add_types
  tsv_fields << :literal if literal

  if fields.any? and not (fields == [:all] or fields == ["all"])
    tsv_fields.concat fields 
  else
    tsv_fields.concat segments.first.annotations if segments.any?
  end
  tsv_fields
  tsv_fields.collect!{|f| f.to_s}
  tsv_fields.delete "offset"
  tsv_fields
end

.sort(segments, inline = true) ⇒ Object

{{{ Sorting



81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
 
# File 'lib/rbbt/ner/segment.rb', line 81

def self.sort(segments, inline = true)
  if inline
    segments.sort do |a,b| 
      case
      when ((a.nil? and b.nil?) or (a.offset.nil? and b.offset.nil?))
        0
      when (a.nil? or a.offset.nil?)
        -1
      when (b.nil? or b.offset.nil?)
        +1
      when (not a.range.include? b.offset and not b.range.include? a.offset)
        a.offset <=> b.offset
      else
        a.segment_length <=> b.segment_length
      end
    end
  else
    segments.sort_by do |segment| segment.offset || 0 end.reverse
  end
end

.split(text, segments, skip_segments = false) ⇒ Object

{{{ Splitting



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
 
# File 'lib/rbbt/ner/segment.rb', line 126

def self.split(text, segments, skip_segments = false)
  sorted_segments = clean_sort segments

  chunks      = []
  segment_end = 0
  text_offset = 0
  sorted_segments.each do |segment|
    return chunks if text.nil? or text.empty?
    next if segment.offset.nil?
    offset = segment.offset - text_offset

    # Consider segment offset. Save pre, or skip if overlap
    case
    when offset < 0 # Overlap, skip
      next
    when offset > 0 # Save pre
      chunk = text[0..offset - 1]
      Segment.setup(chunk, text_offset)
      chunks << chunk
    end

    segment_end = offset + segment.segment_length - 1

    if not skip_segments
      chunk = text[offset..segment_end]
      Segment.setup(chunk, text_offset + offset)
      chunks << chunk
    end

    text_offset += segment_end + 1
    text = text[segment_end + 1..-1]

  end

  if not text.nil? and not text.empty?
    chunk = text.dup
    Segment.setup(chunk, text_offset)
    chunks << chunk
  end

  chunks
end

.tsv(segments, *fields) ⇒ Object 



311
312
313
314
315
316
317
318
319
320
 
# File 'lib/rbbt/ner/segment.rb', line 311

def self.tsv(segments, *fields)
  fields = set_tsv_fields fields, segments
  tsv = TSV.setup({}, :key_field => "ID", :fields => fields, :type => :double)

  segments.each do |segment|
    tsv[segment.id] = self.tsv_values_for_segment(segment, fields)
  end

  tsv
end

.tsv_values_for_segment(segment, fields) ⇒ Object

{{{ Save and load



224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
 
# File 'lib/rbbt/ner/segment.rb', line 224

def self.tsv_values_for_segment(segment, fields)
  info = segment.info
  values = []

  fields.each do |field|
    values << case
    when field == "JSON"
      info.to_json
    when field == "literal"
      segment.gsub(/\n|\t/, ' ')
    when field == "Start"
      segment.offset
    when field == "End"
      segment.end
    else
      info.delete(field.to_sym)
    end
  end

  values
end

Instance Method Details

#end ⇒ Object

{{{ Ranges



22
23
24
25
 
# File 'lib/rbbt/ner/segment.rb', line 22

def end
  return nil if offset.nil?
  offset.to_i + segment_length - 1
end

#includes?(segment) ⇒ Boolean

Returns:

  • (Boolean) 


74
75
76
77
 
# File 'lib/rbbt/ner/segment.rb', line 74

def includes?(segment)
  (segment.offset.to_i >= self.offset.to_i) and
  (segment.offset.to_i + segment.segment_length.to_i <= self.offset.to_i + self.segment_length.to_i)
end

#make_relative(segments, &block) ⇒ Object 



52
53
54
55
56
57
58
59
60
 
# File 'lib/rbbt/ner/segment.rb', line 52

def make_relative(segments, &block)
  if block_given?
    segments.each{|s| s.push offset}
    yield(segments)
    segments.each{|s| s.pull offset}
  else
    segments.each{|s| s.push offset}
  end
end

#offset=(offset) ⇒ Object 



8
9
10
 
# File 'lib/rbbt/ner/segment.rb', line 8

def offset=(offset)
  @offset = offset.nil? ? nil : offset.to_i
end

#pull(offset) ⇒ Object 



32
33
34
35
36
37
38
39
40
 
# File 'lib/rbbt/ner/segment.rb', line 32

def pull(offset)
  if self.offset.nil? or offset.nil?
    self.offset = nil
  else
    self.offset += offset 
  end

  self
end

#push(offset) ⇒ Object 



42
43
44
45
46
47
48
49
50
 
# File 'lib/rbbt/ner/segment.rb', line 42

def push(offset)
  if self.offset.nil? or offset.nil?
    self.offset = nil
  else
    self.offset -= offset 
  end

  self
end

#range ⇒ Object 



27
28
29
30
 
# File 'lib/rbbt/ner/segment.rb', line 27

def range
  raise "No offset specified" if offset.nil?
  (offset.to_i..self.end)
end

#range_in(container = nil) ⇒ Object 



62
63
64
65
66
67
68
69
70
71
72
 
# File 'lib/rbbt/ner/segment.rb', line 62

def range_in(container = nil)
  raise "No offset specified" if offset.nil?
  case
  when (Segment === container and not container.offset.nil?)
    ((offset - container.offset)..(self.end - container.offset))
  when Integer === container
    ((offset - container)..(self.end - container))
  else
    range
  end
end

#segment_length ⇒ Object 



12
13
14
15
16
17
18
 
# File 'lib/rbbt/ner/segment.rb', line 12

def segment_length
  begin
    super()
  rescue
    self.length
  end
end