class Boilerpipe::Filters::DensityRulesClassifier

Public Class Methods

classify(prev, current, nxt) click to toggle source
# File lib/boilerpipe/filters/density_rules_classifier.rb, line 22
def self.classify(prev, current, nxt)
  return false if current.link_density > 0.333333

  if prev.link_density <= 0.555556
    if current.text_density <= 9
      return true if nxt.text_density > 10

      return prev.text_density <= 4 ? false : true
    else
      return nxt.text_density == 0 ? false : true
    end
  else
    return false if nxt.text_density <= 11

    true
  end
end
process(doc) click to toggle source
# File lib/boilerpipe/filters/density_rules_classifier.rb, line 8
def self.process(doc)
  # return doc if doc.text_blocks.size < 2

  empty = Boilerpipe::Document::TextBlock.empty_start
  text_blocks = [empty] + doc.text_blocks + [empty]

  text_blocks.each_cons(3) do |slice|
    prev, current, nxt = *slice
    current.content = classify(prev, current, nxt)
  end

  doc
end