importer.rb.old

require 'rubygems'
require 'open-uri' 
# require 'nokogiri'
uri = 'http://code.google.com/apis/maps/documentation/javascript/reference.html'
doc = open(uri).read()

parts = doc.split(/<h2><a name="[A-Z]\w+"><\/a>/)
first = true
parts.shift()

parts.each do |pt|
  if first : first = false; next; end  
  str = pt
  # print pt

  # name, description, remainder = str.match(/^\s*([A-Z]\w+)<\/h2>.*?<p>(\w+)<\/p>(.*)/m)
  # res = str.match(/^\s*([A-Z]\w+) class<\/h2>[^<]*<p>(.*?)(?=(<\/p>))(.*)/m)
  res = str.match(/^\s*([A-Z]\w+).*?<\/h2>[^<]*<p>(.*?)(?=(<\/p>))(.*)/m)

  # p str

  class_name = res[1]

  class_doc = res[2]
  remains = res[4]
  

  section_parts = remains.split(/<h3>/m)

  section_name = remains.match(/<h3>(.*?)(?=(<\/h3>))/)

  if section_name

    if section_name[1] == 'Constructor'

      res = str.match(/<td>(.*?)(?=(<\/td>))(.*?)<td>(.*?)(?=(<\/td>))/m)
      code = res[1]
      desc = res[4]
      code.gsub!(/<[\w\/ \"\=\#]+>/,"") # extract html content
      code.gsub!(/(\w+)\?/,"?\\1") # swap the question mark side
      ret = ''
    else 

      res = str.match(/<td>(.*?)(?=(<\/td>))(.*?)<td>(.*?)(?=(<\/td>))(.*?)<td>(.*?)(?=(<\/td>))/m)
      p str
      code = res[1]

      ret = res[4]
      desc = res[7]

      code.gsub!(/<[\w\/ \"\=\#]+>/,"") # extract html content
      code.gsub!(/(\w+)\?/,"?\\1") # swap the question mark side

    end
  end

end



# print str


# print str
# print section
# section.scan(/<h3>/){|w| print w}

# parts.each {|p| }