This post originated from an RSS feed registered with Ruby Buzz by Eigen Class. Original Post: Hashes with default procs as memoized functions Feed Title: Eigenclass Feed URL: http://feeds.feedburner.com/eigenclass Feed Description: Ruby stuff --- trying to stay away from triviality. Latest Ruby Buzz Posts Latest Ruby Buzz Posts by Eigen Class Latest Posts From Eigenclass

Thinking of optimizing the small script I wrote long ago to analyze the geographical distribution of eigenclass' readership, I bumped into yet another instance of the "hash with default block as memoized function" idiom. I've been finding that quite often when reviewing older code of mine (or at least feel that way).

That script uses GeoIP's command-line utility to determine which country an IP address is associated to. Calling an external program is really slow so I absolutely want to cache those results:

iptocountry = Hash.new do |h,ip|
  h[ip] = `geoiplookup #{ip}`.chomp.gsub(/^GeoIP Country Edition: /,"")
end
iptocountry.update Marshal.load(File.read("geo.cache")) rescue {}

One minor inconvenience about hashes with default procs is that they cannot be serialized, so it would seem one needs something like

File.open("geo.cache", "w"){|f| Marshal.dump(Hash[*iptocountry.to_a.flatten], f)}

We can do much better though:

iptocountry.default = nil
File.open("geo.cache", "w"){|f| Marshal.dump(iptocountry, f)}

Hash#default= will override the default specified in the Hash.new call.