Forming the base of a tag cloud (See related posts)

a = "curtains in the window I said curtains on the chimney"

# retrieve each word
c = a.split(/\s+/)

# retrieve each unique word
b = c.uniq


#prepare the tags xml document
file_template = File.new('tags_file_template.xml','r')
doc = Document.new(file_template)

#load the tag template
tag_template = doc.root.elements['records/tag'].to_s

b.each do |word|
  tag = Document.new(tag_template)
  tag.root.elements['keyword'].text = word
  tag.root.elements['count'].text = a.scan(/\b#{word}\b/).length
  doc.root.elements['records'].add_element(tag)
end

# count the no of unique words
doc.root.elements['summary/no_of_uniqe_words'].text  = b.length

# get the average no. of words
norm = c.length / b.length.to_f
doc.root.elements['summary/average_word_count'].text  = norm

doc.root.elements.each('//tag') do |node|
  count = node.elements['count'].text.to_f
  weight = 100 / (c.length /  count)
  node.elements['weight'].text = weight
  norm_diff = count - norm
  node.elements['norm_diff'].text = norm_diff > 0 ? '+' + norm_diff.to_s : norm_diff
  node.elements['gauge'].text = (weight / 10).to_i
end

#delete the blank template tag
node = doc.root.elements['records/tag']
node.parent.delete(node)

file_write = File.new('decorating.xml','w')
file_write.puts doc
file_write.close

<tags>
  <summary>
    <no_of_uniqe_words/>
    <average_word_count/>     <!-- the norm -->
  </summary>
  <records>
    <tag>
      <keyword></keyword>
      <count></count>         <!-- e.g. 5 // no of instances of that word -->
      <weight></weight>       <!-- e.g. 50% -->
      <norm_diff></norm_diff> <!-- e.g. +3 // a positive or negative offset from the norm -->
      <gauge></gauge>         <!-- 1 .. 10 -->
    </tag>
  </records>
</tags>