http_log_analyze.rb

def request_reg_exp
  return @_reg_exp if @_reg_exp

  request_type = 'GET'
  request_protocol = '(http:\/\/)?'
  server_name = '([\d\.\w-]+)?'
  server_port = ':?(\d+)?'
  request_address = '([\w\d\.\/]+)'
  request_parameters = '\??(.*)?'
  http_version = 'HTTP\/1\.1'

  reg_exp = '"' +
    request_type + '\s' + request_protocol + server_name + server_port +
    request_address + request_parameters + '\s' + http_version + '"'
  @_reg_exp = Regexp.new(reg_exp, 'i')
end

class Request
  attr_reader :protocol, :hostname, :port, :address, :parameters

  def initialize(matches)
    @protocol = matches[1]
    @hostname = matches[2]
    @port = matches[3]
    @address = matches[4]
    @parameters = matches[5] || ''
  end

  def protocol?
    return @_protocol_ if @_protocol_
    @_protocol_ = self.protocol.blank? 
  end

  def hostname?
    return @_hostname_ if @_hostname_
    @_hostname_ = self.hostname.blank? 
  end

  def port?
    return @_port_ if @_hport_
    @_port_ = self.port.blank? 
  end
end

def analyze_line_with_regexp(log_line)
  matches = request_reg_exp.match(log_line)
  return [] if matches.nil? || matches.length == 1
  request = Request.new(matches)
  return [] if request.protocol? && !request.hostname?
  return [] if !request.protocol? && request.hostname?
  return [] if !request.port? && request.hostname?

  [
    request.address,
    request.parameters
  ]
end