From: "Jesús Gabriel y Galán" <jgabrielygalan@...>
Date: 2012-10-10T23:24:31+09:00
Subject: Re: Simulation of a dictionary

On Tue, Oct 9, 2012 at 9:10 PM, Eric Christopherson
<echristopherson@gmail.com> wrote:
> On Tue, Oct 9, 2012 at 1:17 PM, Joao Silva <lists@ruby-forum.com> wrote:
>> Hi all.
>>
>> What I'm trying to do, is a program that has a hash with predefined
>> words and its corresponding meaning, then when I enter a word that is
>> found in the hash, I must return its meaning, and if not found in the
>> hash, should return the most similar word.
>
> Maybe use something like Levenshtein distance.
>
> Some links to look at:
> http://stackoverflow.com/questions/2158851/grouping-strings-by-similarity
> http://stackoverflow.com/questions/2531502/detect-similar-sounding-words-in-ruby
>

The text gem has an implementation of the Levenshtein distance:

require 'text'

words={
"computer" => "an electronic device ...",
"cat" => "a small domesticated carnivorous ...",
"home" => "the place where one lives permanently ...",
"dog" => "a domesticated carnivorous mammal ... "
}

puts "Enter a word: "
s = gets.chomp

if words.has_key?(s)
  puts words[s]
else
  distances = Hash.new {|h,k| h[k] = []}
  words.keys.each {|word| distances[Text::Levenshtein.distance(word,s)] << word}
  # puts "All words sorted by distance"
  # distances.each {|distance, words| puts "#{distance} => #{words.join(",")}
  puts "nearest words: #{distances[distances.keys.sort.first]}"
end

Jesus.