bwmap.rb

require 'zlib'

APACHE_LINE_RX = /^([0-9\.]+)\s+\S+\s+\S+\s+\[[^\]]+\]\s+\"[^\"]+\"\s+(\d+)\s+(\d+).*$/
def parse_apache_line(text_line)
  if text_line =~ APACHE_LINE_RX
    {:ip => $1, :bytes => $3.to_i}
  else
    nil
  end
end

# Return value is a pair, key-value
def apache_bandwidth_map(text_line)
  fields = parse_apache_line(text_line)
  if fields.nil?
    nil
  else
    {fields[:ip] => fields[:bytes]}
  end
end

def apache_bandwidth_reduce(hash, pair)
  key = pair.keys.first
  old_val = hash[key] || 0
  hash[key] = old_val + pair[key]
  hash
end

def bandwidth_map(file_name_template)
  pairs = []
  Dir.glob(file_name_template) do |fname|
    Zlib::GzipReader.open(fname) do |zstream|
      zstream.each_line do |text_line| 
        pair = apache_bandwidth_map(text_line)
        pairs << pair unless pair.nil?
      end
    end
  end
  reduced_hash = pairs.reduce({}) { |acc, pair| apache_bandwidth_reduce(acc, pair) }
end

## Beginning of main program

ip_to_query = ARGV.first
if ip_to_query.nil?
  STDERR.puts("An IP address is needed as the first argument")
  exit 1
end

bwm = bandwidth_map("*-access.log.gz")
puts bwm[ip_to_query] || "IP not found"