Super fast cache for webservice clients in Ruby

dapatil@scattrbrain.com (Darshan Patil) — Sat, 08 Dec 2007 01:36:00 -0500

I implemented a cache on my website for the AllPosters.com webservice. Instead of caching the raw response from the webservice, I cache the model constructed after parsing the results. The speedup I get on a cache hit is ridiculous because I don’t need to parse data anymore.

My secret weapon was YAML. All I do is take the model constructed after parsing the data and call to_yaml on it. On a cache hit, I simply call YAML::load. Here is the code. Let me know if this can be improved.

require 'digest/sha1'
require 'yaml'

module Utils
  class Cache
    @@cache_dir = "./cache/"

    def self.write(keywords, data)
      File.open(@@cache_dir + digest(keywords), "w") { |f| f.write(data.to_yaml) }
    end

    def self.cached?(keywords)
      File.exists?(@@cache_dir + digest(keywords))
    end

    def self.read(keywords)
      YAML::load(File.open(@@cache_dir + digest(keywords)))
    end

    def self.expire(keywords)
      File.delete(digest(keywords))
    end

    private
    @@lookup = {}
    def self.digest(keywords)
      @@lookup[keywords] ||= Digest::SHA1.hexdigest(keywords)
    end
  end
end

"Super fast cache for webservice clients in Ruby" by Sagen

Fri, 25 Jan 2008 06:57:48 -0500

Keep on blogging, we need you. I’ve got so much useful stuff from your blog and really value you opinion in this stuff.

"Super fast cache for webservice clients in Ruby" by Nensi

Wed, 02 Jan 2008 08:13:06 -0500

Happy holidays to you too. and a belated Christmas wishes..

scattrbrain: Super fast cache for webservice clients in Ruby

Super fast cache for webservice clients in Ruby

"Super fast cache for webservice clients in Ruby" by Sagen

"Super fast cache for webservice clients in Ruby" by Nensi