Parsing bash.org quotes for eggdrop bottalk.tcl

Here is a script to scrape quotes from bash.org, and create a bottalker data file (source’d tcl script) for our eggdrop. If you want to chat with it, come to #oaktoncc on freenode.

require 'rubygems'
require 'open-uri'
require 'hpricot'
require 'cgi'

count = 10
last_page = 407

lines = []

count.times do |i|
  page = last_page - i
  uri = "http://bash.org/?browse&p=#{page}"
  puts "#{uri}"
  doc = Hpricot(open(uri))

  doc.search("//p[@class='qt']").each do |element|

    quote = element.inner_html

    quote.split(/n/).each do |line|
      line = CGI::unescapeHTML(line)
      line.gsub!(//, "")
      line.gsub!(" ", " ")

      # Different formats (otherwise ignore):
      #  blah
      # (@name:#channel) blah
      # name: blah
      # [name] blah

      if line =~ /^s*(.*)/ or line =~ /^s*(.*?)(.*)/ or line =~ /^s*.+?:(.*)/ or line =~ /^s*[.+?](.*)/
        lines << $1.strip
      end

    end

  end
end

File.open("BotTalker_data_bash_org.tcl", "w") do |f|
  f.puts("# Bot Talker Data file.")
  f.puts("set TalkzStrArray {")
  lines.each do |line|
    next if line =~ /[{}\]/ # Ignore lines with {} chars since it screws up the tcl source
    line.gsub!(/[/, "(")
    line.gsub!(/]/, ")")
    f.puts(" {#{line}}")
  end
  f.puts("}")
end