Library to easily parse standard Apache log files.
Switch branches/tags
Nothing to show
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Failed to load latest commit information.

Getting Started

This is a library designed to easily parse and access standard Apache log files.

To get started, require the library:

require 'apache_log_parser'

From there, you would use it as follows:

ApacheLogParser.parse(logfile, rules) do |parsed|
  parsed[:ip]         #=> ""
  parsed[:date]       #=> "21/Jan/2010"
  parsed[:day]        #=> 21
  parsed[:month]      #=> "Jan"
  parsed[:year]       #=> 2010
  parsed[:hour]       #=> 14
  parsed[:zone]       #=> "-0800"
  parsed[:method]     #=> "GET"
  parsed[:http_ver]   #=> "1.1"
  parsed[:resource]   #=> "/some/page.php"
  parsed[:status]     #=> 200
  parsed[:size]       #=> "7047"
  parsed[:referer]    #=> "-"
  parsed[:user_agent] #=> "Mozilla/5.0 (Macintosh; U; Intel..."

The logfile parameter is simply the path to the logfile in question and the rules parameter is a hash of rules to filter the logfile with.

Using Rules

To use rules, simply build a hash with options you want to filter with as follows:

rules = {}
rules[:hour]   = 11..13         # only accept hits between 11:00 and 13:59 hours
rules[:day]    = 21             # only accept hits where the day is 21
rules[:date]   = "12/Jan/2010"  # only accept hits on Jan 12, 2010
rules[:method] = "GET"          # only accept hits where the request method is GET
rules[:status] = 404            # only accept hits where the status response is 404

Rules are inclusive, so only hits where ALL rules are met will be kept. From there, you would simply pass the rules hash into the parse method as shown above. The logfile is read one line at a time, so the memory footprint is quite small and can easily handle large logfiles (sizes as large as 6GB have been tested). The larger the file, the longer the parse process will take though.


Copyright (c) 2012 Joel Watson. See LICENSE for details.