ruby_style_guide_parser.rb

require 'open-uri'
require 'rubocop'
require 'tmpdir'

url = 'https://raw.github.com/bbatsov/ruby-style-guide/master/README.md'
data = open(url).read

code_blocks = data.scan(/```ruby[\r\n]+(.*?)```/mi)
puts "Found #{code_blocks.length} code blocks"

good_blocks = []
# bad_regexp = /\#[^$]*?bad[^$]*?$/
code_blocks.each do |block|
  block = block.first
  if block =~ /\#\s*good/
    # leading code
    match = block.match(/\A(.*?)(?:\#[^$]*?bad[^$]*?$|\#\s*good)/mi)
    good_block = ''
    good_block << match[1] if !match.nil? && match[1].strip.length > 0

    # good blocks
    good_block << block.scan(/(^\s+\#\s*good.*?)(?:\Z|\#[^$]*?bad[^$]*?$)/mi)
                  .flatten.join

    # trailing code
    match2 = block.match(/(?:\#[^$]*?bad[^$]*?$)((?!good)*?)\Z/mi)
    good_block << match2[1] unless match2.nil?

    good_blocks << good_block
  elsif block !~ /\#[^$]*?bad[^$]*?$/ # not good and not bad = good
    # good_blocks << block
    # not always the case.. make the style_guide more consistent
  end
end

good_blocks = good_blocks.join("\n")
good_blocks.gsub!(/^\s*\.+\s*$/mi, '') # remove ... in functions
good_blocks.gsub!(/[ \t]+$/mi, "\n") # remove excess whitespace
good_blocks.gsub!(/\n\n\n+/mi, "\n\n") # remove excess newlines

Dir.mktmpdir do |dir|
  Dir.chdir(dir)
  puts "Writing output to style_guide.rb.."
  open('style_guide.rb', 'w') do |f|
    f << good_blocks
  end

  cli = Rubocop::CLI.new
  cli.run(['style_guide.rb'])
end