insert apache log to mongodb

insert.rb

#!/usr/bin/ruby19
# -*- coding: utf-8 -*-

require 'apachelogregex'
require 'uri'
require 'mongo'

format = '%h %l %u %t \"%r\" %>s %b \"%{Referer}i\" \"%{User-agent}i\" \"%{Cookie}i\"'
parser = ApacheLogRegex.new(format)

datetime_pat = /\A\[(\d{2})\/(\w{3})\/(\d{4}):(\d{2}):(\d{2}):(\d{2}) \+0900\]\z/

connection = Mongo::Connection.new
db = connection.db('apache-log')
logs = db.collection('log')

discarded_lines = 0

ARGF.each do |line|
  unless log =  parser.parse(line.chomp)
    discarded_lines += 1
    next
  end
  next unless log['%>s'].to_s == '200'
  next unless log['%r'] =~ /^GET/i

  matched = datetime_pat.match(log['%t'])
  next unless matched

  time = Time.local(matched[3], matched[2], matched[1], matched[4], matched[5], matched[6])
  bytes = log['%b'].to_i

  begin
    uri = URI.parse(URI.escape(log['%r'].sub(/^GET /, 'http:/'))).normalize.to_s
  rescue
    next
  end

  h = { :time => time, :bytes => bytes, :uri => uri }
  logs.insert(h)
end

STDERR.puts "discarded lines: #{discarded_lines}"
STDERR.puts 'done'