Convert HTML to markdown using pandoc

convert.rb

#!/usr/bin/env ruby
#

%w[nokogiri open-uri chronic].each{|lib| require lib}

base  = "https://blog.tilton.co"
doc   = Nokogiri::HTML open "#{base}/archive.html"

doc.search('section.archives a').each do |post|
  post_link     = post[:href]
  post_title    = post.text
  post_tags_wc  = post.next_element.text.split(';')[0]
  post_tags     = post_tags_wc.split(/,\s+/) if not post_tags_wc.nil?
  post_basename = File.basename(post_link, File.extname(post_link))

  cmd = "pandoc -s -r html #{base}/#{post_link} -o ./scrape/#{post_basename}.md"
  markdown    = system(cmd)
  infile        = "./scrape/#{post_basename}.md"

  re1 = "\n\nPlease enable JavaScript to view the..*[email protected]\\).\n"
  re2 = "---.*{#post-title}\n----+$\n\n"
  File.open(infile) do |f|
    f = f.read.gsub(Regexp.new(re1, Regexp::MULTILINE), '')
    f = f.gsub(Regexp.new(re2, Regexp::MULTILINE), '')
    f = f.gsub(/\n\n(\#\#\# edited at )(.*)( {\#post-edited})$/, '')

    f.gsub(/^.*(\#\#\# )(.*)( {.*{\#post-created})$/, '')
    post_created = Date.strptime("#{$2} 12:00:00", '%a %b %d, %Y').strftime('%Y-%m-%d %H:%M:%S')

    f = f.gsub(/.*{\#post-created}.*\n\n/, '')

    post_slug = post_basename.split('-')[3..-1].join('-')

    outfile = "_posts/#{post_basename}.md"
    puts outfile
    File.open(outfile, "w") do |f2|
      f2.write(<<EOF
created:            #{post_created}
edited:
title:              #{post_title}
slug:               #{post_slug}
draft:              False
tags:               #{post_tags.join("\n") if not post_tags.nil?}

#{f}


<!--
vim: ft=markdown
-->
EOF
)
    end

  end
end